US8160093B2 - Timing system for modular cable modem termination system - Google Patents
Timing system for modular cable modem termination system Download PDFInfo
- Publication number
- US8160093B2 US8160093B2 US12/413,690 US41369009A US8160093B2 US 8160093 B2 US8160093 B2 US 8160093B2 US 41369009 A US41369009 A US 41369009A US 8160093 B2 US8160093 B2 US 8160093B2
- Authority
- US
- United States
- Prior art keywords
- docsis
- shelf
- timestamp
- packet
- phy
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active, expires
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
- H04N21/239—Interfacing the upstream path of the transmission network, e.g. prioritizing client content requests
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/0706—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment
- G06F11/0709—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment in a distributed system consisting of a plurality of standalone computer nodes, e.g. clusters, client-server systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/60—Network structure or processes for video distribution between server and client or between remote clients; Control signalling between clients, server and network components; Transmission of management data between server and client, e.g. sending from server to client commands for recording incoming content stream; Communication details between server and client
- H04N21/61—Network physical structure; Signal processing
- H04N21/6106—Network physical structure; Signal processing specially adapted to the downstream path of the transmission network
- H04N21/6118—Network physical structure; Signal processing specially adapted to the downstream path of the transmission network involving cable transmission, e.g. using a cable modem
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/60—Network structure or processes for video distribution between server and client or between remote clients; Control signalling between clients, server and network components; Transmission of management data between server and client, e.g. sending from server to client commands for recording incoming content stream; Communication details between server and client
- H04N21/61—Network physical structure; Signal processing
- H04N21/6156—Network physical structure; Signal processing specially adapted to the upstream path of the transmission network
- H04N21/6168—Network physical structure; Signal processing specially adapted to the upstream path of the transmission network involving cable transmission, e.g. using a cable modem
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/16—Analogue secrecy systems; Analogue subscription systems
- H04N7/173—Analogue secrecy systems; Analogue subscription systems with two-way working, e.g. subscriber sending a programme selection signal
- H04N7/17309—Transmission or handling of upstream communications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04H—BROADCAST COMMUNICATION
- H04H20/00—Arrangements for broadcast or for distribution combined with broadcast
- H04H20/65—Arrangements characterised by transmission systems for broadcast
- H04H20/76—Wired systems
- H04H20/77—Wired systems using carrier waves
- H04H20/78—CATV [Community Antenna Television] systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04J—MULTIPLEX COMMUNICATION
- H04J3/00—Time-division multiplex systems
- H04J3/02—Details
- H04J3/06—Synchronising arrangements
- H04J3/0635—Clock or time synchronisation in a network
- H04J3/0638—Clock or time synchronisation among nodes; Internode synchronisation
- H04J3/0658—Clock or time synchronisation among packet nodes
- H04J3/0661—Clock or time synchronisation among packet nodes using timestamps
- H04J3/0667—Bidirectional timestamps, e.g. NTP or PTP for compensation of clock drift and for compensation of propagation delays
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L7/00—Arrangements for synchronising receiver with transmitter
- H04L7/0008—Synchronisation information channels, e.g. clock distribution lines
Definitions
- CMTS Cable Modem Termination System
- DOCSIS Data Over Cable Service Interface Specification
- CMTS complementary metal-oxide-semiconductor
- MAC DOCSIS Media Access Control
- the conventional CMTS architecture does not scale well since any one of the separate components in the CMTS can limit processing capacity and only a limited number of DOCSIS packet processing devices and physical interfaces can be located in the same CMTS chassis.
- CMTS complementary metal-oxide-semiconductor
- the present invention addresses this and other problems associated with the prior art.
- a modular Cable Modem Termination System includes a packet shelf operating a Data Over Cable Service Interface Specifications (DOCSIS) Media Access Control (MAC) framer.
- DOCSIS Data Over Cable Service Interface Specifications
- MAC Media Access Control
- PHY Physical Interface
- PHY Physical Interface
- One or more upstream PHY shelves send DOCSIS data received from an upstream path of the cable plant over the packet switched network to the packet shelf.
- Hybrid Fiber Coax (HFC) plant may be replaced with different PHY components for other access technologies such as wireless, Digital Subscriber Lines (DSL), Ethernet-to-the-Home, Fiber-to-the-Home, or fiber Passive Optical Networks (PONs).
- DSL Digital Subscriber Lines
- Ethernet-to-the-Home Ethernet-to-the-Home
- Fiber-to-the-Home Fiber Passive Optical Networks
- FIG. 1A is a diagram of a modular Cable Modem Termination System (CMTS).
- CMTS Cable Modem Termination System
- FIG. 1B is a more detailed diagram of the modular CMTS shown in FIG. 1A .
- FIG. 2 shows an alternative embodiment of the modular CMTS.
- FIG. 3 shows another alternative embodiment of the modular CMTS.
- FIG. 4 shows yet another alternative embodiment of the modular CMTS.
- FIG. 5 shows the different data streams sent and received by a remote Physical Interface (PHY) in the modular CMTS.
- PHY Physical Interface
- FIGS. 6A and 6B show how the modular CMTS generates a tunnel.
- FIG. 7 shows how two devices conduct an Internet Protocol (IP) session through the modular CMTS.
- IP Internet Protocol
- FIGS. 8A-8C show packet formats used by the modular CMTS.
- FIG. 8D shows a more detailed diagram of the downstream remote PHY used in the modular CMTS.
- FIG. 9A shows a DOCSIS request-grant protocol.
- FIG. 9B shows how DOCSIS maps are transported over a packet switched network in one embodiment of the modular CMTS.
- FIG. 10 shows how the modular CMTS performs an early map release.
- FIG. 11 shows how the modular CMTS establishes DOCSIS tunnels for different Quality of Service (QoS) levels.
- QoS Quality of Service
- FIG. 12 shows how one embodiment of the modular CMTS handles packet latency conditions.
- FIG. 13 shows how another embodiment of the modular CMTS handles packet latency conditions.
- FIG. 14 is a more detailed block diagram of an upstream PHY shelf.
- FIG. 15 is a diagram of upstream data block.
- FIG. 16 is a diagram of a domain header sent along with the upstream data blocks.
- FIG. 17 is a diagram of a timestamp snapshot block.
- FIG. 18 is a diagram of a DOCSIS packet.
- FIG. 19 shows how Quality of Service (QoS) values are associated with Service Identifiers (SIDs).
- FIG. 20 shows a MAP packet used for assigning QoS values to Service Identifiers.
- FIG. 21 is a block diagram showing how QoS values are assigned to DOCSIS packets.
- FIG. 22 is a block diagram showing how the upstream PHY shelf performs in a best effort mode.
- FIG. 23 is a block diagram showing how the upstream PHY shelf conducts early request extraction.
- FIG. 24 shows another embodiment of early request extraction.
- FIG. 25 shows the control messages sent from the packet shelf to the remote PHY shelves.
- FIG. 26 shows how a MAP advance time is used in a cable network.
- FIG. 27 shows one embodiment of how a timing system is used in the modular CMTS.
- FIGS. 28A-28D are diagrams showing different hardwired timing system configurations.
- FIG. 29 is a diagram showing a star wired timing system.
- FIG. 30 is a diagram showing a daisy chained timing system.
- FIG. 31 shows a packet shelf in the modular CMTS operating as a timestamp master.
- FIG. 33 shows how DOCSIS ranging is implemented in the modular CMTS.
- FIG. 34 shows a measurement packet used during DOCSIS ranging.
- FIGS. 35 and 36 show how timestamps are de-jittered in the modular CMTS.
- MAC Media Access Control Used to refer to the layer 2 element of the system which would include DOCSIS framing and signaling.
- PHY Physical Layer Used to refer to the downstream QAM transmitters and the upstream burst demodulators.
- PHY SHELF In general, a line card with DOCSIS PHYs and located in a switch or router. QAM Quadrature Amplitude Modulation or Quadrature Amplitude Modulator UDP User Datagram Protocol US Upstream
- FIG. 1A shows a cable network 10 that includes a modular Cable Modem Termination System (CMTS) 14 .
- the modular CMTS 14 includes a packet shelf 16 that communicates with a remote cable Physical Interface (PHY) shelf 18 over a packet switched network 26 .
- the packet switched network 26 is a Gigabit Ethernet (GE) network.
- GE Gigabit Ethernet
- the GE network 26 is only one example and any type of packet switched network can alternatively be used to connect the packet shelf 16 to the remote PHY shelf 18 .
- the packet shelf 16 and PHY shelf 18 can be implemented using packet processing routers or switches. Of course, other network processing devices can also be used.
- the packet shelf 16 includes a Data Over Cable Service Interface Specifications (DOCSIS) packet processor 20 that operates a DOCSIS Media Access Controller (MAC) 22 .
- DOCSIS Data Over Cable Service Interface Specifications
- MAC DOCSIS Media Access Controller
- a GE interface port 24 is used by the DOCSIS MAC 22 to communicate with the PHY shelf 18 over network 26 .
- a router or routing processor 13 is located either internally or externally with the packet shelf 16 . The router 13 transfers packets between the packet shelf 16 and a Wide Area Network (WAN) 12 .
- WAN Wide Area Network
- the remote PHY 18 includes one or more separate downstream PHY shelves 30 , one or more separate upstream PHY shelves 32 , and possibly one or more timing shelves 28 . It is also possible to locate the timing shelf 28 with the packet shelf 16 or include the timing shelf with the upstream PHY shelf 30 or downstream PHY shelf 32 . Any combination of chassis can be used to house the different shelves 28 , 30 and 32 . In one example, each chassis contains at least one downstream PHY shelf 30 and one upstream PHY shelf 32 . In an alternative embodiment, different chassis may contain one or more downstream PHY shelves 30 or one or more upstream PHY shelves 32 , but not both. In another embodiment, the remote PHY shelf 18 may contain one or more downstream PHY shelves, while the upstream PHY 32 is combined with the rest of the CMTS in a conventional CMTS chassis.
- the PHY shelf 18 is connected to one or more Hybrid Fiber/Coax (HFC) plants 34 that each include a downstream path 40 and an upstream path 42 .
- the HFC 34 is a broadband bidirectional shared-media transmission system that uses fiber trunks between the PHY shelf 18 and fiber nodes (not shown). Coaxial distribution is provided from the fiber nodes to customer locations 38 .
- the endpoints 38 of the cable network 10 are alternatively referred to as Customer Premise Equipment (CPE) and can include any wired or wireless device that needs to communicate over WAN 12 .
- CPE 38 may include any type of computer server, laptop, Personal Computer (PC), etc that communicates over the HFC 34 through a Cable Modem (CM) 36 .
- the Cable Modem 36 may be located in the CPE 38 , may be located in a separate chassis, or may be integrated into a Set Top Box (STB) (not shown).
- STB Set Top Box
- the cable modem 36 operates a DOCSIS MAC that conducts DOCSIS messaging and transfers DOCSIS frames with the DOCSIS MAC 22 in packet shelf 16 . Operation of the cable modem 36 is known to those skilled in the art and is therefore not described in further detail.
- the modular CMTS 14 decouples the backplane communications that were previously required between a CMTS DOCSIS MAC and the CMTS PHY interface that are used for communicating over cable plant 34 .
- This allows the DOCSIS MAC 22 in packet shelf 16 to communicate to the PHY shelf 18 remotely over packet switched network 26 and allows the downstream PHY shelf 30 to operate independently from the upstream PHY shelf 32 .
- the modular CMTS components can be more effectively matched with different cable network requirements. For example, if more cable modems 36 are connected to the HFC 34 , more downstream PHY shelves 30 and/or upstream PHY shelves 32 can be added to the modular CMTS 14 to support the increased DOCSIS bandwidth demand.
- FIG. 1B shows the modular CMTS 14 in more detail.
- Each packet shelf 16 can include multiple DOCSIS MAC cards 22 that each communicate over the packet network 26 through GE ports 24 .
- the downstream PHY shelf 30 includes an Ethernet port 30 A that sends DOCSIS data 30 H to a QAM modulator 30 F.
- Native elementary stream video encapsulated in UDP or RTP and received on Ethernet port 30 A is processed by a video processor 30 B before being sent to the QAM 30 F.
- Timestamp information is sent to a timestamp de-jitter element 30 C and then rewritten in a timestamp rewrite element 30 D before being sent to the QAM 30 F.
- the modulated output signal from the QAM 30 F is passed through an up-converter 30 G before being sent on the downstream path 40 of the HFC 34 .
- the upstream PHY shelf 32 includes an Ethernet port 32 A that both receives DOCSIS messages from the packet shelf 16 and sends DOCSIS messages and data to packet shelf 16 .
- a QAM demodulator 32 C demodulates signals received on the upstream path 42 of the HFC 34 .
- a DOCSIS Remote PHY Interface (DRPI) framer 32 B frames the data received over upstream path 42 for transport over the packet switched network 26 .
- DRPI DOCSIS Remote PHY Interface
- FIG. 2 shows an alternative embodiment of the modular CMTS 14 where the packet shelf 16 also includes a DOCSIS line card 44 that connects to the HFC 34 .
- the PHY shelf 18 still includes the timing shelf 28 , downstream shelf 30 and upstream shelf 32 .
- the DOCSIS line card 44 in the packet shelf 16 may handle all upstream DOCSIS traffic and the remote PHY 18 may only include downstream PHY shelf 30 for handling downstream DOCSIS traffic.
- the remote PHY 18 may still include upstream PHY shelves 32 , but fewer than the remote PHY shelf 18 shown in FIG. 1 .
- FIG. 3 shows another alternative embodiment of the modular CMTS 14 where multiple packet shelves 16 connect to multiple downstream PHY shelves 30 A- 30 N and also connects to multiple upstream PHY shelves 32 A- 32 N.
- Each downstream PHY shelf 30 connects to the downstream path 40 of the same or different HFC 34 and each upstream PHY shelf 32 connects to the upstream path 42 of the same or different HFC 34 .
- Any combination of upstream PHY shelves 30 and downstream PHY shelves 32 can be contained in the same chassis.
- FIG. 4 shows another embodiment where two different packet shelves 16 _ 1 and 16 _ 2 are each connected to the same PHY shelves 18 _ 1 and 18 _ 2 .
- packet shelf 16 _ 1 and packet shelf 16 _ 2 may each connect to the same downstream PHY shelves 30 _ 1 and 30 _ 2 and the same upstream PHY shelves 32 _ 1 and 32 _ 2 .
- the configuration in FIG. 4 allows for additional DOCSIS MAC capacity by having the two packet shelves 16 _ 1 and 16 _ 2 communicate with either or both PHY shelf 18 _ 1 or 18 _ 2 .
- one of the packet shelves 16 _ 1 or 16 _ 2 can operate as a backup DOCSIS MAC, if the primary packet shelf is disabled or fails.
- the PHY shelves 18 _ 1 and 18 _ 2 also receive video data 50 from one or more video servers 48 over packet switched network 26 .
- the video server 48 sends video data 50 to the downstream PHY shelves 30 _ 1 and 30 _ 2 for modulation over the downstream path 40 of the different HFCs 34 .
- downstream PHY shelf 30 receives both Moving Picture Experts Group Transport Stream (MPEG-TS) video data 50 from the video server 48 shown in FIG. 4 and also receive the DOCSIS downstream data 54 from the packet shelf 16 ( FIG. 4 ).
- the upstream shelf 32 sends DOCSIS upstream data 56 back to the packet shelf 16 and also generates spectrum management data 58 that may be used by the packet shelf 16 , or some other management server, to monitor signaling on the HFC 34 .
- a Digital Signal Processor (DSP) 60 in the upstream shelf 32 may generate analysis data from the frequency spectrum 62 of the upstream path 42 .
- the upstream shelf 32 then converts the analysis data into packets that are then sent over the packet switched network 26 to the packet shelf 16 or other management server.
- DSP Digital Signal Processor
- a remote PHY downstream protocol determines how DOCSIS messages and data is transferred between the MAC 22 in the packet shelf 16 and the downstream PHY shelf 30 and how messaging is sent from the MAC 22 to the upstream PHY shelf 32 .
- the remote PHY downstream protocol is unidirectional and no acknowledgement messages are required in an upstream direction.
- data messages are not acknowledged but control messages are acknowledged.
- FIG. 6A shows the functional elements in packet shelf 16 in more detail.
- Packets 70 are received from a routing device that is either internal or external to the packet shelf 16 .
- Some operations 72 such as a rate shaping, accounting, Quality of Service (QoS), etc. are performed on the incoming packets 70 .
- the packets 70 are converted into DOCSIS frames 84 by a DOCSIS framer 74 and then converted into MPEG frames 86 by an MPEG framer 76 that are alternatively referred to as MPEG packets.
- the DOCSIS framer 84 generates DOCSIS frames 84 from packets 70 received over WAN 12 ( FIG. 1 ) or generates DOCSIS frames 84 for DOCSIS messages, such as for MAP 80 .
- An upstream scheduler 78 generates DOCSIS MAPs 80 , for example, in response to DOCSIS transmit requests 82 received from cable modems 36 ( FIG. 1A ).
- the MAPs 80 identify timeslots allocated to the requesting cable modem 36 for transmitting data in the upstream path 42 of HFC plant 34 ( FIG. 1A ).
- the MAPs 80 are sent to the MPEG framer 76 and formatted into MPEG frames 86 that are sent to the Ethernet framer 77 along with the MEPG frames 86 containing packet data 70 .
- the MEPG framer 86 is used in one embodiment of the modular HFC 14 .
- a Packet Streaming Protocol (PSP) is used for transporting DOCSIS frames from the packet shelf 16 to the PHY shelf 18 .
- PSP Packet Streaming Protocol
- DOCSIS frames 84 are concatenated together into a continuous data stream 94 .
- the concatenated data stream 94 is then broken up into separate packet fragments 98 that each include a PSP header 96 .
- the PSP headers 96 may include pointers 97 that identify where different DOCSIS frames 84 are located.
- the concatenation 94 can lower the Packet Per Second (PPS) throughput in switching engines used in the modular CMTS system.
- PPS Packet Per Second
- the concatenated data stream 94 is then fragmented to meet the Maximum Transfer Unit (MTU) of the packet switched network media 26 ( FIG. 1A ).
- MTU Maximum Transfer Unit
- the DOCSIS frames 84 generated by the DOCSIS framer 74 are not formatted into MPEG frames 86 and alternatively sent directly to the Ethernet framer 77 .
- the conversion of DOCSIS frames to MPEG frames is performed in the PHY shelf 18 .
- the packet shelf 16 creates a tunnel 88 over the packet switched network 26 for transporting the DOCSIS frames to the downstream PHY shelf 30 .
- an MPEG tunnel 88 is used.
- a DOCSIS tunnel is used. If a DOCSIS tunnel is used, a new field 85 may be used in the DOCSIS header for packet fragmentation or concatenation. This allows DOCSIS frames to be fragmented over multiple Ethernet packets.
- the fragmentation or concatenation field 85 contains numbers identifying the sequence for the fragmented DOCSIS frames 84 .
- the Ethernet framer 77 encapsulates one or more MPEG frames 86 into the payload of the same Ethernet frame 89 that includes an Ethernet header 90 and a Cyclic Redundancy Check (CRC) 92 .
- CRC Cyclic Redundancy Check
- FIG. 7 shows a logical representation of the tunnel 88 formed in the packet switched network 26 between the packet shelf 16 and the PHY shelf 18 .
- Examples of MPEG and DOCSIS tunnels 88 are referred to below in FIGS. 8A-8D .
- any type of tunneling protocol or Virtual Private Network (VPN) connection can be used over the packet switched network 26 .
- VPN Virtual Private Network
- L2TP Layer-two Tunneling Protocol
- L2TPV3 Layer-Two Tunneling Protocol Version 3
- PPTP Point-to-Point Tunneling Protocol
- the L2TP protocol is described in U.S. Pat. No. 5,918,019, issued Jun.
- Layer-3 tunnels can alternatively be used.
- the tunnel 88 can alternatively be established over a non-Ethernet interface using for example, Resilient Packet Ring (RPR).
- RPR Resilient Packet Ring
- An IP connection 102 is established between a device 100 on the WAN 12 and a client or CPE 38 in the cable network 10 through tunnel 88 .
- the tunnel 88 allows the packet shelf 16 to perform the DOCSIS MAC operations and then transport the DOCSIS frames over the packet switched network 26 to the PHY shelf 18 .
- the downstream PHY shelf 30 is therefore not required to perform DOCSIS MAC operations 22 ( FIG. 1A ).
- FIG. 8A shows one example of a UDP transport packet format used for encapsulating and transporting the DOCSIS frames 84 over the packet switched network 26 to the PHY shelf 18 .
- UDP transport packet format
- MPT MPEG Transport Stream
- PSP Packet Streaming Protocol
- source shelf refers to the packet shelf 16 and the term “destination shelf” refers to the downstream PHY shelf 30 .
- An Ethernet header 110 is defined by IEEE-802.3. Upon transmission of Ethernet packet 89 by the source shelf, the Ethernet destination address 112 will be the Ethernet address of the destination shelf or of the next hop router in packet switch network 26 .
- the Ethernet source address 113 Upon reception of frame 89 by the destination shelf, the Ethernet source address 113 will be the Ethernet address of the output port of the source shelf or of the previous hop router in network 26 .
- An Ethernet 802.1Q header 114 is defined by IEEE-802.1Q. Header 114 is optional and provides frame prioritization and Virtual Local Area Network (VLAN) support at layer 2 of the Open System Interconnect (OSI) model. In one implementation when header 114 is used, the Length/Type field 115 of the 802.3 header 110 is set to an appropriate 802.1Q TagType value.
- the packet shelf 16 and the upstream PHY shelf 32 optionally may or may not support header 114 .
- An IP header 116 is defined by RFC-791.
- An IP source address 118 in header 116 is the IP address of the source shelf.
- An IP destination address 120 is the IP address of the destination shelf.
- the IP destination address 120 may be an IP unicast or IP multicast address.
- the IP header 116 uses IPv4 or IPv6. Of course any version of the IP protocol can be used.
- the Quality of Service (QoS) information associated with the tunnel 88 is contained in a DiffServ (DS) field 119 of the IP header 116 .
- DS DiffServ
- a UDP header 122 is defined by RFC-768 and the CRC 132 is defined by IEEE-802.3.
- GRE Generic Routing Encapsulation
- the UDP source port 124 is unused and is set by the source shelf to all zeros. The destination shelf may ignore the UDP source port 124 .
- the UDP destination port 126 is a common agreed upon value between the source shelf and the destination shelf.
- the source shelf might set the UDP checksum 128 to zero and the destination shelf may have the ability to ignore this UDP checksum field.
- the payload of UDP packet 89 for the MPT mode is shown in FIG. 8B .
- the downstream PHY shelf 30 accepts one to seven (or more for systems with an MTU larger than 1500 bytes) MPEG-TS packets 86 within the UDP payload 130 ( FIG. 8A ) of the UDP packet 89 .
- the packet shelf 16 in one embodiment does not put stuffing bytes between the UDP header 122 and the first MPEG-TS header or between consecutive MPEG-TS packets 86 .
- the maximum length of 7 MPEG-TS packets 86 are allowed within the UDP/IPv4 packet within a 802.3 Ethernet frame with 802.1Q tagging. Of course other Ethernet frame lengths can also be used and the maximize number of packets allowed can be varied.
- the downstream PHY shelf 30 may generate NULL MPEG-TS frames when there are no MPEG-TS packets 86 to be transmitted.
- the payload 130 of a UDP packet 89 for PSP mode is shown in FIG. 8C .
- the PSP Payload Data Unit (PDU) 134 C consists of one or more DOCSIS frames.
- the DOCSIS frame includes a fully formatted DOCSIS header, any extended headers, the DOCSIS payload, and the CRC.
- the Packet Streaming Protocol (PSP) can take a series of layer 3 packets, layer 2 frames, or any fixed or variable sized pieces of data, assemble them as a stream of back to back Payload Data Units (PDUs) 134 D, and then break that stream up into segments and encapsulate each segment with a PSP header 134 B. In doing so, the first and last PDU 134 D of a PSP packet 89 may be fragmented.
- the PSP has two basic modes for recovering individual PDU frames 134 out of a series of PSP packets 134 .
- the two modes are interpretive and non-interpretive.
- the mode to be used, and the PDU type (for example, DOCSIS, IP, Ethernet), are negotiated as part of the assignment of the payload type.
- the PSP packet 134 has one PDU pointer.
- the PDU pointer points to the first PDU 134 D that has its first byte in the PSP packet 134 . If the PDU pointer is non-zero, the bytes prior to the location pointed to by the first pointer are considered a fragmented PDU 134 D, and are combined with the last PDU of the previous PSP packet 134 .
- the receiver then starts to interpret the contents of the PDU based upon the protocol type associated with the payload type. When it reaches the end of the DOCSIS frame, but it has not reached the end of the PSP packet 134 , it assumes another PDU 134 D is appended, and begins to interpret that PDU. By doing so, all PDUs 134 D are broken apart and reassembled.
- the PSP packet 134 has one PDU pointer for each start of a PDU 134 D.
- the same PDU fragment rule applies if the first SOP is non-zero. Since each PDU 134 D is explicitly called out, the receiver that manages the reassembly of the PDUs 134 D does not have to be aware of the protocol within the PDU. If the last PDU pointer does not have the End bit E set, then the PDU 134 D is considered a fragment and is held until the next PSP packet 134 is parsed.
- FIG. 8D shows a simplified block diagram of the internal data paths of the downstream PHY shelf 30 previously shown in FIG. 1B .
- the Downstream PHY shelf 30 may receive MPEG elementary streams which have been encapsulated in MPT packets 86 ( FIG. 8B ) and placed in a UDP datagram. This is referred to as the video MPT flow 30 K.
- a transparent MPT flow 30 H is received and transmitted to the QAM interface 30 F without any interpretation of the contents of those MPT packets.
- There may be more than one transparent MPT flow 30 H where each flow has a different Differentiated Services Code Point (DSCP) and different addressing, but is destined for the same QAM channel.
- DSCP Differentiated Services Code Point
- All DOCSIS frames including packet based frames and MAC management based frames, are included within the MPT flow 30 J.
- the downstream PHY shelf 30 searches the MPT payload for any DOCSIS SYNC messages and performs SYNC corrections 30 T. It then forwards the MPT packet to the QAM interface 30 F. In the MPT mode, MPT frames can be received by the downstream PHY shelf 30 and forwarded directly to the modulation interface 30 F without having to terminate and regenerate the MPT framing. Except for manipulation of the payload of the MPT frames for DOCSIS in the SYNC correction 30 T.
- the Packet Streaming Protocol is a layer 3 convergence layer protocol which allows packets to be consecutively streamed together and fragmented at arbitrary boundaries.
- the intent of the PSP mode is to facilitate Quality of Service (QoS).
- QoS Quality of Service
- This mode is used for transporting traditional DOCSIS data and signaling messages which use one or more DSCP values. For example, in order to reduce REQ-GNT latency, MAP MAC management messages may be sent using a different DSCP on a different PSP flow 30 N than the rest of the DOCSIS channel.
- Each PSP flow 30 N is received, terminated, and the DOCSIS frames within the flow are extracted by PSP termination 30 W.
- the DOCSIS frames from all the combined PSP flows 30 N are sorted into output queues 30 P by DSCP Mapping 30 V based upon the DSCP value contained within the IP packet in the DOCSIS payload.
- the outputs of the QoS queues 30 P go to a packet scheduler 30 Q which decides which queue 30 P is to be serviced.
- the packet scheduler 30 Q is also responsible for inserting DOCSIS SYNC messages 30 R within the time interval specified by DOCSIS timing 30 C.
- the output of the packet scheduler 30 Q goes to a MPT Transmission Convergence (TC) engine 30 S that places the DOCSIS frames into MPT frames.
- TC Transmission Convergence
- the video MPT flow 30 K, the transparent MPT flow 30 L, the DOCSIS MPT flow 30 M, and the PSP flow 30 N provide four MPT flows into the MPT scheduler 30 E.
- the MPT scheduler 30 E arbitrates between the four streams and makes the decision which MPT packet will be transmitted at what time.
- the MPT scheduler 30 E receives its scheduling policies from an Edge Resource Management Interface (ERMI) (not shown) and takes into account the DSCP values used within and across the various types of flows.
- ERMI Edge Resource Management Interface
- the destination IP address 120 ( FIG. 8A ) of the packets 89 is used to choose a downstream PHY shelf 30 .
- the IP address selects a PHY chassis and a PHY chassis may have one or more IP addresses.
- the destination UDP port 126 is used by the packet shelf 16 and the downstream PHY shelf 30 to select a downstream QAM 30 F. More than one destination UDP port 126 may point to the same downstream QAM 30 F. Packets destined to different QAMs 30 F are addressed to different UDP destination ports 126 if they are on the same IP subnet.
- the DOCSIS frames contain IP packets which have a Differentiated Services Code Point (DSCP).
- the Type of Service (TOS) bits are a subset of the DSCP bits.
- the DSCP is a value located in the DiffServ field 119 of the IP header 116 ( FIG. 8A ) and is used by routers to determine the correct Per Hop Behaviour (PHB).
- the DSCP 119 may be used at the egress of the packet shelf 16 , within the network 26 from the packet shelf 16 to the downstream PHY shelf 30 , and at the ingress of the downstream PHY shelf 30 . Examples of traffic that may require flows with different DSCP are signaling, such as DOCSIS MAPs, SYNCs, UCDs; DOCSIS MAC management messages; packet cable signaling messages; data; VoIP; and video.
- each unique flow may have a different DSCP 119 .
- the PDU DSCP may be defined as the DSCP of the IP Packet contained with the DOCSIS frame contained within the PDU 134 D. Different PDU DSCPs may be mapped into different PSP flows 30 N where each PSP flow 30 N would have a different DSCP 119 . More than one PDU DSCP may map to the same PSP flow 30 N.
- Each unique PSP stream 30 N is assigned a unique destination UDP port 126 .
- the packet switched network 26 between the packet shelf 16 and the downstream PHY shelf 30 may have a certain Maximum Transfer Unit (MTU).
- MTU Maximum Transfer Unit
- the MTU for an Ethernet network might be 1522 bytes.
- One technique for determining this value is to have both endpoints run MTU path discovery. If a maximum size DOCSIS frame were to be tunneled from the packet shelf 16 to the downstream PHY shelf 30 , the MTU of the resulting packet would be greater than 1522 bytes. Both the MPT and PSP modes avoid this issue by offering streaming which concatenates and fragments packets.
- FIG. 9A shows one example of the DOCSIS protocol used for transferring data from the cable modem 36 to the packet shelf 16 .
- Data is sent by the cable modem 36 in timeslots that are allocated by the DOCSIS MAC 22 .
- a MAC management message (MAP) is sent from the packet shelf 16 that allocates the timeslot transmission opportunities to the cable modem 36 in response to a request message 136 .
- MAP MAC management message
- the cable modem 36 sends the data transmit request message 136 through the upstream path 42 of the HFC 34 to the upstream PHY shelf 32 .
- the upstream PHY shelf 32 forwards the request 136 over packet switched network 26 to the DOCSIS MAC 22 in packet shelf 16 .
- the MAC 22 responds with a grant MAP 137 that travels back over the packet switched network 26 to the downstream PHY shelf 30 .
- the downstream PHY shelf 30 forwards the grant MAP 137 over the downstream path 40 of HFC 34 to the cable modem 36 .
- the grant MAP 137 identifies a timeslot in the future allocated to cable modem 36 for transmitting data.
- the cable modem 36 transmits data 138 over upstream path 42 of HFC 34 during the timeslot allocated in grant MAP 137 .
- the upstream PHY shelf 32 forwards the data 138 to the packet shelf 16 over network 26 .
- the time required to conduct this ‘request-grant-transmit data’ exchange creates a delay when sending data from the cable modem 36 to another device in the WAN 12 (FIG. 1 A). This delay can be further aggravated by the additional delay that may happen transporting the messages 136 and 137 and the data 138 over the packet switched network 26 . The delay problem can be even further aggravated by the tunneling scheme described above in FIGS. 6-8 .
- the request message 136 is received by the upstream scheduler 78 over the packet switched network 26 .
- the upstream scheduler 78 generates the grant MAP 137 as described above in FIG. 9A .
- the MAP 137 is formatted into a DOCSIS frame which is then formatted into an MPEG packet by MPEG framer 76 .
- Ethernet framer 77 then combines the MPEG packets containing the MAP 137 with other MPEG packets that may contain other DOCSIS messages, DOCSIS data, or video data into the same payload 130 ( FIG. 8A ) of Ethernet packet 89 .
- Ethernet framer 77 may encapsulate up to seven MPEG packets 138 into the same Ethernet packet 89 .
- the Ethernet packets 89 are also rate shaped for transmission at a predetermined rate. This combining of MPEG packets and rate shaping further delays the transmission of MAP 137 to the cable modem 36 .
- FIG. 10 shows an early MAP release scheme used for reducing delays in sending MAP 137 to the cable modem 36 ( FIG. 9A ).
- MAP 137 has been formatted into an MPEG-TS packet 138 by MPEG framer 76 and encapsulated into the UDP payload 130 ( FIG. 8A )
- the packet shelf 16 stops accumulating more MPEG-TS packets 138 into the UDP payload 130 , and transmits (releases) the Ethernet packet 144 .
- Other Ethernet packets 142 that do not contain MAP messages or other time sensitive DOCSIS messages accumulate the normal number of MPEG packets 138 .
- Ethernet packets 144 containing DOCSIS frames 84 may be given higher priority than Ethernet packets 142 that do not contain DOCSIS frames.
- Ethernet packets 142 may only contain MPEG-TS packets 138 with video data and may accordingly be given lower priority than Ethernet packet 144 that contains DOCSIS MAP message 137 .
- the packet shelf 16 can contain a high QoS queue 150 and a low QoS queue 152 as described below in FIG. 11 .
- the MPEG-TS packets 138 containing DOCSIS frames may be loaded into a high QoS queue 150 and the MPEG-TS packets 138 that do not contain DOCSIS frames may be loaded into the low QoS queue 152 .
- the Ethernet packets are output from the high QoS queue 150 with higher priority than the Ethernet packets output from low QoS queue 152 .
- packets 154 - 160 with different Quality of Service (QoS) requirements may be received by the packet shelf 16 .
- QoS Quality of Service
- video packets 154 , data packets 156 , voice packets 158 and signaling packets 160 each may have different associated QoS requirements.
- the different packets 154 - 160 each have associated Type Of Service (TOS) values 155 in their IP headers that correspond with the required QoS.
- the TOS values 155 are alternatively referred to above as Differentiated Services Code Point (DSCP) values.
- DSCP Differentiated Services Code Point
- the MPEG encapsulation described above in FIGS. 6-10 may combine multiple MPEG or DOCSIS packets into the same tunnel.
- a best effort tunnel 162 is used for transporting the different packets 154 - 160 over the packet switched network 26 .
- the multiple packets 154 - 160 are encapsulated into the same Ethernet packet 163 with no consideration of the associated TOS values 155 or other QoS values. Therefore, packets with different TOS values 155 are encapsulated and sent by the packet shelf 16 over the packet switched network 26 in the same tunnel 162 .
- the packet shelf 16 includes a processor 20 and multiple packet queues 150 and 152 that are used for storing packets 154 - 160 according to their corresponding QoS values.
- the QoS corresponds to the packet TOS values 155 .
- the queue 150 is associated with high QoS packets and the queue 152 is associated with lower QoS packets.
- the DOCSIS packet processor 20 receives and loads the different packets 154 - 160 into either high QoS queue 150 or low QoS queue 152 according to their corresponding TOS values 155 .
- packets with a TOS value 155 above a predetermined threshold value are loaded into high QoS queue 150 and any packets with a TOS value 155 below the predetermined threshold value are loaded into low QoS queue 152 .
- the voice packet 154 and several signaling packets 160 have TOS values 155 above the predetermined threshold and are accordingly loaded into high QoS queue 150 .
- the data packets 156 and the video packets 154 have TOS values below the predetermined QoS threshold and are accordingly loaded into low QoS queue 152 .
- the processor 20 then generates separate high and low QoS tunnels 164 and 166 corresponding to the packets in queues 150 and 152 .
- the voice packet 158 and the signaling packets 160 are encapsulated and transported through the packet switched network 26 over high QoS tunnel 164 .
- the data packets 156 and the video packets 154 in low QoS queue 152 are encapsulated and transported through the network 26 over low QoS tunnel 166 .
- the number of MPEG or DOCSIS frames that are combined in the different tunnels 164 and 166 may vary according to priority.
- the high QoS tunnel 164 may encapsulate fewer MPEG or DOCSIS frames together than the lower QoS tunnel 166 .
- the packet shelf 16 provides different QoS tunnels to the remote PHYs 18 .
- Any number of different QoS tunnels can be provided according to system requirements. For example, separate tunnels may be established for each different type of packet data 154 - 160 .
- the processor 20 can also establish the different tunnels according to parameters other than, or in combination with, the TOS value 155 .
- the tunnels 164 and 166 can also be established according to different Packet Identifier (PID) values used for identifying different MPEG streams.
- PID Packet Identifier
- the different tunnels 164 and 166 may be established according to different User Datagram Protocol (UDP) port values identified in the UDP header 122 ( FIG. 8A ).
- the tunnels may also be established according to any combination of the TOS, PID, and UDP port values.
- the tunnels can be established according to any combination of UDP, IP or MPEG parameters or according to the different types of data associated with the packets 154 - 160 .
- FIG. 12 shows how the modular CMTS 14 addresses latency conditions.
- packets may be delayed while being transmitted over the network 26 .
- the QAMs 30 F FIG. 1B
- the downstream PHY shelves 30 may insert MPEG nulls into the downstream path 40 of the HFC 34 .
- the late arriving packets can build up in the output queue 176 .
- Packet shelf 16 To prevent packet buildup in queue 176 , the packet shelf 16 rate shapes packet traffic to the downstream PHY shelf 30 .
- Packet shelf 16 includes a packet queue 172 that receives packets 170 .
- the downstream PHY shelf 30 includes an input packet queue 174 and the output packet queue 176 .
- the output packet queue 176 has a payload rate of 100% of the QAM bandwidth.
- the payload rate of the packet queue 172 in the packet shelf 16 is set to some value less than the 100% of the payload rate for packet queue 176 .
- the packet shelf 16 will then deliver packets at a Variable Bit Rate (VBR) or Constant Bit Rate (CBR) that is less than the maximum payload rate for the output queue 176 . This allows the downstream PHY shelf 30 to empty output queue 176 even after a latency period where MPEG nulls have been inserted.
- VBR Variable Bit Rate
- CBR Constant Bit Rate
- the processor 20 may selectively vary the payload rate capacity of buffer 172 according to an amount of jitter detected in packet switched network 26 . For example, a high jitter condition may be detected in packet switched network 26 by the processor 20 using existing DOCSIS ranging operations. The processor 20 then accordingly may reduce the payload rate capacity of packet queue 172 . This further reduces the chance of packets backing up in the output queue 176 .
- FIG. 13 shows an alternative embodiment of a latency management system where the packet shelf 16 includes a high QoS queue 178 and a low QoS queue 180 .
- the downstream PHY shelf 30 has a corresponding high QoS queue 184 and a low QoS queue 186 that are coupled between the input queue 174 and the output queue 176 .
- the packets in queues 178 and 180 can in combination have 100% of the payload rate of the output queue 172 .
- Packets sent from high QoS queue 178 are output by the packet shelf 16 with a higher priority than the packets in the low QoS queue 186 .
- the packets from high QoS queue 178 received by the downstream PHY shelf 30 are stored in high QoS queue 184 and the packets from low QoS queue 180 received by the downstream PHY shelf 30 are stored in low QoS queue 186 .
- the packets in low QoS queue 186 can be delayed or even dropped when packets backup in output queue 176 due to latency conditions.
- FIG. 14 shows one embodiment of the upstream PHY shelf 32 in more detail that includes multiple cable Physical Interface (PHY) elements each containing QAMs 32 C that demodulate signals received over the upstream path 32 of a HFC plant 34 .
- the PHY interface that contains QAMs 32 C is referred to generally as PHY 32 C.
- the upstream PHY shelf 32 uses a similar packet format to that previously shown in FIG. 8A to transport DOCSIS MAC PHY Interface (DMPI) blocks 200 in packets 203 over the packet switched network 26 to the packet shelf 16 .
- DMPI is described in the DOCSIS 2.0 Radio Frequency Interface Specification which is herein incorporated by reference.
- the packets 203 sent from the upstream PHY shelf 32 to the packet shelf 16 containing the DMPI blocks 200 are referred to generally as DMPI over IP (DoIP) packets 203 .
- the upstream PHY protocol used for transporting the DoIP packets 203 includes both control plane and forwarding plane modes.
- the upstream PHY shelf receives DMPI blocks from the PHY, assembles them into DOCSIS packets, and forwards those DOCSIS packets to the M-CMTS using the same packet format that is used in the downstream direction.
- the upstream PHY shelf 32 takes content off the HFC plant 34 and sends it over the network 26 to the packet shelf 16 .
- Cable modems 36 ( FIG. 1A ) send upstream DOCSIS packets 202 in the form of Forward Error Correction (FEC) blocks mapped to a QAM payload that are then demodulated by the QAM receivers in PHYs 32 C and converted into DMPI blocks 200 .
- FEC Forward Error Correction
- the DOCSIS REMOTE PHY Interface (DRPI) framer 32 B encapsulates the DMPI blocks 200 into Ethernet frames as described above in FIG. 8A and then sends the DoIP packets 203 over the packet switched network 26 to packet shelf 16 .
- the packet switched network 26 is an Ethernet network.
- any IP based network may be used.
- the upstream PHY shelf 32 may encapsulate multiple DMPI blocks 200 received on the same upstream paths 42 into the same DoIP packets 203 or may alternatively encapsulate DMPI blocks 200 generated from different upstream paths 42 A- 42 C into the same DoIP packets 203 .
- FIG. 15 shows the UDP payload 204 that is contained within the DoIP packet 203 shown in FIG. 14 .
- the headers in the DoIP packet 203 are similar to those shown above in FIG. 8A .
- a header 206 of the payload 204 contains a packet sequence number in header 206 that increases by one for each transmitted packet 203 . When a maximum number is reached, the sequence number in header 206 begins again from zero. The sequence number in header 206 allows the packet shelf 16 to reassemble the DMPI blocks 200 in a correct sequence after being transmitted over network 26 .
- a single DoIP packet 203 may contain DMPI blocks 200 from more than one upstream path 42 and more than one logical channel on the same upstream path 42 .
- the UDP destination port 126 ( FIG. 8A ) represents a group of one or more physical upstream channels.
- the upstream PHY shelf 32 in one embodiment does not put stuffing bytes between the UDP header 112 ( FIG. 8A ) and the header in the first DMPI block 200 or between consecutive headers of the DMPI blocks 200 .
- the format of the DMPI blocks 200 is defined in the DOCSIS 2.0 specification and is therefore not described in further detail.
- the supported block types are FIRST_DATA block, MIDDLE_DATA block, LAST_DATA block, PHY_STATUS block, and NO_BURST block.
- the CHANNEL block is discarded and a DOMAIN block 208 is included as described below in FIG. 16 .
- the DMPI domain block 208 is a new block definition and may not be supported by a conventional native DMPI interface on the upstream PHY chip.
- the conventional DMPI specification defined in DOCSIS 2.0 is per channel.
- the DMPI domain block 208 allows DMPI blocks 200 from multiple channels 42 A- 42 C ( FIG. 14 ) to be multiplexed together into one DoIP packet 203 , and provides a way for the packet shelf 16 to sort incoming intermixed DMPI blocks 200 into correct upstream channel queues.
- the DMPI domain block 208 is inserted once at the beginning of a Type-Length-Value (TLV) section of the UDP payload 204 ( FIG. 15 ).
- the DMPI domain block 208 is inserted for example when a next DMPI data block 200 is from a different logical or physical upstream channel 42 than the previous DMPI block 200 .
- the DMPI domain block 208 may contain a domain ID that identifies the particular PHY interface 32 C and HFC 34 associated with the DMPI block 200 and an Upstream Channel ID (UCID) corresponds to the unique logical channel ID that the DOCSIS data 202 is received over. Thus, every unique logical upstream channel connected to every PHY 32 C ( FIG. 14 ) is uniquely identified.
- UID Upstream Channel ID
- FIG. 17 shows an Upstream Channel Descriptor (UCD) snapshot of an upstream frame counter 212 , mini-slot counter 214 , and timestamp 216 published by the packet shelf 16 .
- This is referred to generally as the timestamp snapshot 210 .
- the timestamp 216 is a counter that all cable network components use as a reference.
- the mini-slot number 214 is used for identifying when cable modems 36 ( FIG. 1A ) are allowed to transmit data over the upstream path 42 of HFC 34 .
- the packet shelf 16 publishes the timestamp snapshot 210 , it is created by the upstream PHY shelf 32 by the PHY 32 C.
- the upstream PHY shelf 32 then periodically sends the snapshot 210 to the packet shelf 16 .
- the packet shelf 16 maintains a copy of the frame counter 212 and mini-slot counter 214 for each logical upstream channel which is time aligned with its copy of the timestamp counter 216 . These three counters are used to create the timestamp snapshot 210 in an Upstream Channel Descriptor (UCD) message 364 ( FIG. 25 ).
- the packet shelf can alternatively use the snapshot directly from the PHY shelf.
- the upstream PHY shelf 32 then either periodically sends a copy of its timestamp snapshot 210 to the packet shelf 16 or sends a copy during different conditions, such as during initialization of the upstream channel or when the UCD changes.
- the packet shelf 16 time aligns the frame counter 212 and the mini-slot counter 214 from the upstream PHY self timestamp snapshot 210 to the packet shelf timestamp counter and stores those values as the current frame counter and mini-slot counter values.
- An “open” DMPI data block 200 sequence is when a first block has been received but a last block has not been received (if PHY_STATUS is not expected) or if an expected PHY_STATUS has not been received.
- an open sequence may be closed by reception of the LAST_DATA block and the optional PHY_STATUS block. For example, if the PHY_STATUS block is being used, a block sequence is not completely closed until the PHY_STATUS block is received by the packet shelf 16 .
- the packet shelf 16 receives out-of-sequence DoIP packets 203 , it waits a certain amount of time to determine if the packet or packets have been misordered (i.e. the missing packets are received eventually) or lost (they are not received within the timeout period). If the packet shelf 16 determines that one or more DoIP packets 203 have been lost, the DMPI blocks 200 from all open block sequences are dropped until these open sequences can be terminated. Termination occurs by the reception of new FIRST_DATA block from the same PHYs 32 C and logical channels as the block sequences affected by the dropped packets 203 .
- DOCSIS packets 202 are received by the upstream PHY shelf 32 .
- the DOCSIS packets 202 include a DOCSIS header 224 , an extended header field 226 , Ethernet header 228 , IP header 218 , an application dependant UDP header 230 , a tunnel header such as RTP or L2TPv3 232 , payload 234 , and CRC 236 .
- the DOCSIS packets 202 can have different QoS priorities. For example, different QoS values 222 may be contained within the IP header 218 received by the upstream PHY shelf 32 .
- a portion 220 of the DOCSIS packet 202 may be encrypted by the cable modem 36 .
- some of the header information that includes the QoS priority information 222 may be suppressed in the DOCSIS packet 202 .
- the upstream PHY shelf 32 may not be able to identify the QoS priority information 222 .
- the challenge then is how to maintain priority for the DOCSIS packets 202 when they are sent over the packet switched network 26 to the packet shelf 16 .
- FIG. 19 shows two techniques that can be used for maintaining or providing QoS for DoIP packets 203 sent from the upstream PHY shelf 32 to the packet shelf 16 .
- the upstream Print shelf 32 maintains a table 252 that associates Service Identifiers (SIDs) 253 with different Type of Service (TOS) or DSCP values 255 .
- SIDs Service Identifiers
- TOS Type of Service
- the SIDs 253 are normally assigned by the packet shelf 16 to active or admitted upstream service flows 42 on the HFC 34 .
- the packet shelf 16 can also identify TOS values 255 associated with the SID values 253 .
- the upstream packet shelf 32 inserts the SID-TOS values into table 252 and uses table 252 to then prioritize the DOCSIS data 202 .
- the DOCSIS priority can be used as an index to a DSCP/ToS code point.
- the QoS values can be assigned by the packet shelf 16 according to different IP addresses, types of data, UDP port assignments, etc. associated with the cable modems 36 .
- the software in the packet shelf 16 sends the SID-TOS mappings 254 to the upstream PHY shelf 32 for all the different SID values 253 that may be used by PHY 32 C ( FIG. 14 ).
- the SID-TOS mappings 254 may be sent during initialization, registration, or dynamic service setup, and stored in table 252 .
- the upstream PHY shelf 32 uses the table 252 to assign the DOCSIS packets 202 priorities as described below.
- the CMTS forces a ToS corresponds to a DOCSIS priority on the packets sent from the cable modem.
- the packet shelf 16 dynamically sends different SID-TOS values 255 to the upstream PHY shelf 32 in DRPI MAPs 256 .
- FIG. 20 shows the DRPI MAP 256 in more detail.
- the MAP 256 is conventionally used for identifying mini-slots for cable modem upstream transmissions and includes a conventional DRPI header 258 , SID value 260 , mini-slot value 263 , and mini-slot length value 264 .
- the MAP 256 now also includes new SID to TOS mapping 266 that dynamically associates a TOS value to the SID values 253 for DOCSIS packets received on the HFC 34 .
- the PHY 32 C in the upstream PHY shelf 32 uses the SID value 260 and mini-slot values 262 to demodulate data bursts from cable modems.
- the DRPI framer 32 B then uses the SID-TOS mapping 266 associated with the demodulated DMPI blocks 200 ( FIG. 14 ) to prioritize the DoIP packets 203 sent over network 26 ( FIG. 19 ).
- the SID-TOS MAPs 256 allows the DRPI framer 32 B ( FIG. 14 ) to only keep track of a relatively small number of SID-TOS values 266 . Further, if a TOS value 255 for a particular SID 253 is dynamically changed during transmission of MAP 256 , there are no synchronization problems that arise with other upstream PHY shelves 32 . There may be no need to send a new mapping each time a modem is registered or a dynamic service is set up.
- FIG. 21 shows DOCSIS packets 202 received on the upstream path 42 of the HFC 34 by the PHY 32 C.
- the DOCSIS packets 202 have associated SID values 270 .
- the PHY 32 C converts the received DOCSIS frames 202 into DMPI blocks 200 .
- the DRPI framer 32 B uses the SID values 270 associated with the DMPI blocks 200 to identify associated TOS values 255 in table 252 and loads the DMPI blocks 200 into the high, medium, or low QoS queues 272 , 274 or 276 , respectively, associated with the identified TOS values 255 .
- the DMPI blocks 200 associated with SID values 1 and 4 have high TOS values and are accordingly loaded by the DRPI framer 32 B into the high QoS queue 272 .
- the DRPI framer 32 B encapsulates the DMPI blocks 200 from high QoS queue 272 into a DoIP packet 278 assigned a high QoS value.
- a high TOS or DSCP value 282 is assigned to the DS field 119 in the IP header 116 ( FIG. 8A ) of DoIP packet 278 .
- the DMPI blocks 200 having SID values 270 associated with low TOS values 255 are loaded by PHY 32 C into the low QoS queue 276 .
- the DRPI framer 32 B then encapsulates the DMPI blocks 200 in low QoS queue 276 into a low QoS DoIP packet 280 having a low TOS or DSCP value 282 .
- the upstream PHY shelf 32 uses a best effort mode where DMPI blocks 200 are encapsulated into DoIP packets 292 without any consideration for associated TOS values. However, request messages may still be sent in separate high priority DoIP packets as described below in FIG. 23 .
- DOCSIS frames 202 can be sent by the cable modems 36 ( FIG. 1A ) in many different formats and can contain different types of information.
- the DOCSIS frames can contain data 308 or bandwidth request (REQ) messages 311 .
- the cable modems 36 send REQ messages 311 to the packet shelf 16 to request data transmission in the upstream path 42 of the HFC 34 .
- This transmission REQ message 311 may be sent in a separate DOCSIS REQ frame 310 or may be piggybacked along with data in DOCSIS frame 312 .
- the REQ message 311 is sent along with multiple concatenated data packets in DOCSIS frame 314 . It is also possible that the REQ message 311 may be embedded in one of the data concatenated data packets in DOCSIS frame 314 .
- the REQ message 311 is combined with fragmented DOCSIS frames 316 A- 316 C.
- the PHY 32 C converts the DOCSIS frames 308 - 316 into DMPI blocks 200 .
- the DRPI framer 32 B then encapsulates multiple DMPI blocks 200 together into DoIP packets that are then transported over the network 26 .
- a REQ message 311 may be the first DMPI block 200 encapsulated in a DoIP packet that contains multiple DMPI blocks 200 .
- the REQ message 311 could be therefore be delayed until several other DOCSIS frames 202 are received, converted into DMPI blocks 200 , and then encapsulated into the same DoIP packet.
- the upstream PHY shelf 32 conducts an early REQ extraction.
- the PHY 32 C monitors the bytes of the incoming DOCSIS frames 202 for REQ messages 311 . Whenever a REQ message 311 is detected, the PHY 32 C generates a separate DMPI REQ block 318 that is sent to a separate REQ message queue 306 . As soon as the DMPI REQ block 318 is received in REQ queue 306 , the DRPI framer 32 B formats the REQ block 318 into a DoIP packet 320 and sends it over the network 26 to the packet shelf 16 . This eliminates the possible delay that could be created encapsulating the DMPI REQ block 318 with other DMPI blocks 200 .
- the DoIP REQ packet 320 is assigned a high priority QoS or DSCP value.
- the DRPI framer 32 B assigns a high TOS value 322 to DoIP REQ packet 320 . This allows packet 320 to be processed with higher priority through the packet switched network 26 ( FIG. 1A ).
- the REQ message 311 may be replicated in DMPI REQ block 318 and the original REQ message 311 processed in a normal manner.
- the REQ message 311 is also converted into DMPI blocks 200 , loaded into a data queue 304 in the DRPI framer 32 B along with other DOCSIS data, and then encapsulated with other DOCSIS data 326 into DoIP packet 324 .
- the DoIP packet 324 may or may not be assigned a particular TOS value 328 based on the priority criteria discussed above in FIGS. 21 and 22 .
- the same REQ message 311 may be sent in both the DoIP REQ packet 320 and the DoIP packet 324 .
- the packet shelf 16 may process whichever REQ message 311 is received first in packet 320 or 324 .
- the PHY 32 C may not be able to extract and separately transmit a DoIP REQ packet 320 for all received REQ messages 311 .
- the PHY 32 C may send DRPI framer 32 B an indication whenever a REQ message 311 is successfully detected.
- the DRPI framer may then insert an indicator in a designated field in either or both of the DoIP REQ packet 320 and DoIP packet 324 that notifies the packet shelf 16 that the same REQ message 311 has been sent in two different DoIP packets.
- the packet shelf 16 then ignores the second received REQ message 311 .
- FIG. 24 shows another embodiment of a multi-channel upstream PHY shelf 32 .
- the multiple PHYs receivers 32 C_ 1 - 32 C_ 3 convert the DOCSIS frames 202 received over the different upstream channels 42 A- 42 C into DMPI blocks 200 A- 200 C, respectively.
- the DRPI framer 32 B receives the DMPI blocks 200 A- 200 C into different queues 340 A- 340 C and may combine the DMPI blocks 200 A- 200 C associated with the different upstream paths 42 A- 42 C, respectively, into the same DoIP packets 344 . This reduces latency by not having to wait and use multiple DOCSIS frames from a same upstream path 42 for creating DoIP packet 344 .
- the upstream PHY shelf 32 can also use the configuration in FIG. 24 in combination with another embodiment of the early REQ release.
- the REQ messages 311 are encapsulated along with other DMPI blocks 200 A- 200 C by the DRPI framer 32 B into the same DoIP packet 346 .
- the DoIP packet 346 is released as soon as the DMPI block 200 containing the REQ message 311 is encapsulated.
- the next DoIP packet 348 then starts encapsulating DMPI blocks 200 where DoIP packet 346 left off.
- the DoIP packet 346 containing the REQ message 311 may optionally be tagged with a high TOS value 350 by the DRPI framer 32 B. If other REQ messages 311 are detected from other upstream paths 42 A- 42 C, they may be combined with the REQ message 311 already loaded into DoIP packet 346 .
- FIG. 25 shows the signaling that is transmitted between the packet shelf 16 , upstream PHY shelf 32 , and downstream PHY shelf 30 .
- the DOCSIS packets 202 are received by the upstream PHY receiver 32 C and converted into DoIP packets 366 by the DRPI framer 32 B and sent from the upstream PHY shelf 32 to the packet shelf 16 .
- the DOCSIS MAC 20 in the packet shelf 16 programs the downstream PHY shelf 30 and upstream PHY shelf 32 by sending MAP messages 360 , SYNC messages 362 and Upstream Channel Descriptor (UCD) messages 364 .
- the MAP messages 360 are used for allocating timeslots as described above.
- the SYNC messages 362 are used for timing and the UCD messages 364 contain programming parameters for the PHY shelves 30 and 32 .
- the messages 360 , 362 and 364 can be encapsulated into tunnels as described above.
- the specific contents of the MAP, SYNC, and UCD messages are described in the DOCSIS 2.0 specification and are therefore not described in further detail.
- the contents of the MAP, UCD, and SYNC messages can be sent with the control messages of a tunneling protocol.
- the MAP, UCD, and SYNC messages could be represented in data fields in Attribute Value Pairs (AVPs) with the Layer 2 Tunneling Protocol version 3 (L2TPv3).
- AVPs Attribute Value Pairs
- L2TPv3 Layer 2 Tunneling Protocol version 3
- the REQ message 136 can represent either a DOCSIS request Information Element (IE) that is successfully sent through a contention request slot, or a piggyback request.
- the GNT 137 is used for the explicit scheduling of an upstream packet and is typically sent in the DOCSIS MAP message as shown in FIG. 9B .
- FIG. 26 shows a REQ-GNT flow for a conventional non-modular cable system.
- a CMTS 398 schedules a REQ opportunity 402 in a MAP 400 A. This could be an explicit contention REQ IE, or it could be a data transmit IE on which the REQ 402 is piggybacked with other data. Starting with this as a reference point, the REQ opportunity 402 propagates down the HFC plant creating delay 404 .
- a cable modem 399 sends back a REQ 406 which has delay 408 created by the both the HFC plant and the ranging delay in the cable modem 399 . Ranging works by adding delay to cable modems so that they all appear to be at the worst case plant delay.
- the REQ 406 with delay 408 is received by a non-modular CMTS 398 , it goes through the physical interface (PHY), an input queuing process, and finally is processed by an upstream scheduler 78 shown in FIG. 6 .
- the upstream scheduler 78 may decide to delay the packet for QoS reasons such as rate shaping or because it has decided that other REQs have higher priority at that moment.
- the upstream scheduler 78 creates a MAP 400 for placing a GNT 412 .
- the upstream scheduler 78 creates a MAP message 400 B that represents some time in the future.
- the time and associated MAP message 400 B is chosen using a multi-step process. First, a parameter called the MAP advance time 414 is added to a current time reference (timestamp). Then, a further addition of time is added to get to the next MAP boundary 400 B. Then the packet data Information Element (IE) is scheduled within that MAP boundary 400 B. The cycle then continues.
- IE packet data Information Element
- the MAP advance time 414 is used by the CMTS 398 to allow for a variety of time delays in the round trip path both internal and external between the CMTS 398 and the cable modem 399 .
- One of the values used in the MAP advance time 414 is derived either statically or dynamically and takes into account the delay required to send signals over the HFC plant 34 ( FIG. 1A ) and provides a design margin referred to as the “safety” factor.
- the MAP advance time 414 takes into account the HFC plant round trip time, cable modem 399 minimum processing delay, CMTS 398 receive queue delay, CMTS 398 transmit queue delay, CMTS 398 transmit PHY delay (interleaver), and some margin.
- the modified REQ-GNT data path for the remote PHY system 10 shown in FIGS. 1A and 9A has additional delays associated with the processing time of the downstream remote PHY 30 , upstream remote PHY 32 , MAC 22 , and latency in the packet switched network 26 .
- the timing and synchronization system described below provides DOCSIS timing and synchronization for the remote PHY architecture 10 and compensates for the additional delays created in the modular CMTS.
- FIG. 27 shows timing shelves 28 that are used for aligning/synchronizing the frequency (clock) and DOCSIS timestamp (TS) for the different remote PHY elements, including the packet shelf 16 , downstream PHY shelf 30 and upstream PHY shelf 32 .
- clock frequency
- TS DOCSIS timestamp
- One technique described below uses a hardwired timing system and a second technique sends a timestamp over the packet switched network 26 .
- FIG. 28A shows a hardwired timing configuration.
- Separate timing shelves 28 are directly connected to the packet shelf 16 and to the PHY shelves 30 and 32 .
- the timing shelves 28 include a frequency master module 422 that generates a clock signal 423 and a time stamp master module 424 that generates timestamp 425 .
- the packet shelf 16 , downstream PHY shelves 30 , and the upstream PHY shelves 32 all operate a frequency slave 426 that runs off clock 423 and a timestamp slave 428 that runs off the timestamp 425 .
- the timing shelves 28 generate the common clock signal 423 and timestamp reference 425 and distribute that information electrically over point to point connections 432 (twisted pair or coax transmission) to each shelf (or line card if need be) in the cable system.
- the same timing shelf 28 might be used for all the different CMTS shelves. If any combination of the packet shelf 16 , downstream PHY shelf 30 , or upstream PHY shelf 32 are located remotely from each other, then separate timing shelves 28 may be used at each remote location.
- Global positioning system (GPS) receivers 430 are connected to each timing shelf 430 and provide a frequency reference and clock used for synchronizing the different autonomously operating timing shelves 28 .
- FIG. 28B shows an alternative embodiment where the frequency master 422 and the timestamp master 424 operate in the packet shelf 16 and are hardwired directly to frequency slaves 426 and timestamp slaves 428 , respectively, in the downstream PHY shelf 30 and upstream PHY shelf 32 .
- FIG. 28C shows another embodiment where the frequency master 422 and the timestamp master 424 operate in the downstream PHY shelf 30 and drive the frequency slaves 426 and timestamp slaves 428 in the packet shelf 16 and upstream PHY shelf 32 through direct hardwired connections.
- the upstream PHY shelf 32 operates the frequency master 422 and timestamp master 424 .
- the packet shelf 16 and the downstream PHY shelf 30 contain frequency slaves 426 and timestamp slaves 428 that are driven by direct hardwired connections to the frequency master 422 and timestamp master 424 in the upstream PHY shelf 32 .
- FIG. 29 shows a star wired configuration used for connecting redundant timing shelves 28 and 28 _B to different remote shelves.
- Each block 432 represents a packet shelf 16 , downstream PHY shelf 30 , or upstream PHY shelf 32 .
- the frequency slave 426 and the timestamp slave 428 in each shelf 432 is connected to the frequency master 422 and the timestamp master 424 in the timing shelf 28 . If the main timing shelf 28 fails, a backup timing shelf 28 _B has an independently operating frequency master 422 and timestamp master 424 that can each be independently connected to the frequency slaves 426 and timestamp slaves 428 , respectively, in each shelf 432 .
- FIG. 30 shows an alternative embodiment where the frequency and timestamp connections 433 are wired to the shelves 434 , 435 , 436 , and 437 in a daisy chain configuration.
- the primary timing shelf 28 connects to the backup timing shelf 28 _B and to two shelves 434 and 436 .
- the backup timing shelf 28 _B connects to the two shelves 434 and 436 .
- Shelves 434 and 436 then connect the clock signal 423 and timestamp 425 to shelves 435 and 437 .
- multiple different shelves 434 - 437 are all located in a same chassis.
- a common backplane or network is used for connecting the clock signal 423 and the timestamp signal 425 from the timing shelf 28 to shelves 434 - 437 . It is also possibly that one or more of the timing shelves 28 or shelves 434 - 437 forward the timing signals onto additional timing shelves 28 .
- Sending a DOCSIS timestamp is one embodiment. Sending a timestamp value that is not in DOCSIS format is also possible. The main distinction is that the timestamp is sent in a software message over a network rather than over a dedicated timing interface.
- the timing system in FIG. 31 separates frequency and timestamp generation. It is assumed that the packet shelves 16 , and PHY shelves 30 and 32 , can run stand-alone. In this timing configuration, there may be hardwired point to point connections between different frequency circuits 422 , but the timestamp circuits 424 and 428 typically communicate via the packet switched network 26 .
- a DOCSIS timestamp is sent.
- a timestamp value is sent that is not in the DOCSIS format. The distinction is that the timestamp is sent in a software message over a network rather than over a dedicated timing interface.
- each shelf 16 , 30 and 32 may generate its own internal clock frequency using a frequency master 422 . These internal frequency references each drive an internal timestamp counter.
- a timing shelf 28 operating a frequency master module 422 may supply a hardwired clock signal 423 to frequency slaves 426 similar to that shown in FIG. 28 .
- a frequency module in one of the shelves 16 , 30 or 32 may operate as a frequency master 422 and then supply a hardwired clock signal 423 to the other packet shelves 16 or PHY shelves 30 and 32 in the same physical location.
- the packet switched network 26 eliminates some of the physical limitations of conventional CMTS systems.
- timing and clock signals have to be either star-wired or daisy chained between multiple CMTS shelves in the same chassis.
- the star-wired connectively requires separate ports for each slave component and the daisy chained connectivity can cause the clock to degrade when passing through each CMTS component.
- the packet switched connectivity shown in FIG. 31 only requires a single network connection in the timestamp master 424 and timestamp slaves 428 for exchanging timestamp information.
- each shelf 16 , 30 and 32 may be slightly different and the timestamp values for each system component may drift from each other.
- one of the timestamp generators operates as the timestamp master 424 .
- the timestamp generator in the packet shelf 16 is the timestamp master 424 and the other timestamp counters in the downstream PHY shelf 30 and the upstream PHY shelf 32 operate as timestamp slaves 428 .
- the timestamp slaves 428 continuously receive timestamps 430 over the IP network 26 from the timestamp master 424 , dejitter them, and then update their timestamp counters accordingly.
- one of the downstream PHY shelves 30 or upstream PHY shelves 32 may operate as the timestamp master 424 .
- the downstream PHY shelf 30 overwrites the timestamp value 430 in the DOCSIS stream with a corrected and de-jittered timestamp. This is referred to as DOCSIS timestamp correction and is described in more detail below in FIGS. 35 and 36 .
- a primary MAC 22 operates as the timestamp master 424 and a backup MAC 22 _B operates as a timestamp slave 428 . If the primary operating MAC 22 fails, the backup MAC 22 _B is brought on line without disrupting the timestamp value. For example, whenever the timestamp slave 428 in MAC 22 _B does not receive the timestamp 430 after some period of time, it may automatically convert over to operating as the timestamp master 424 .
- the backup timestamp master operation provided by MAC 22 _B can alternatively be implemented in one of the PHY shelves 30 or 32 .
- the timestamp 430 is sent by MAC 22 to the PHY shelves 30 and 32 in a DOCSIS SYNC message 362 (see FIG. 25 ).
- the SYNC message 362 can be embedded within the data stream 88 that runs between the MAC card 22 and the downstream and upstream PHY shelves 30 and 32 as shown in FIGS. 6-8 .
- the data stream 88 may use the same packet format used in video on demand (VOD) systems for sending content to edge QAMs.
- VOD video on demand
- a UDP packet 89 shown in FIG. 8A may be used for carrying the timestamp 430 to the PHY shelves 30 and 32 .
- the UDP packet 89 may contain multiple MPEG-TS frames 86 .
- the MAC 22 may place the SYNC message 362 in the same position each time within the data stream 88 .
- the MAC 22 may place the SYNC message 362 in a last MPEG-TS frame 86 in the UDP packet 89 shown in FIG. 8A .
- the SYNC message 362 may be sent over the packet switched network 26 in separate unicast packets having IP addresses associated with the different packet or PHY shelves 30 and 32 .
- the MAC 22 may send a single SYNC message 362 in a multicast packet that is received by each of the packet and PHY shelves 30 and 32 .
- the timestamp slaves circuits 428 in the upstream PHY shelf 32 and in the backup MAC 16 _B may be simpler than the timestamp slave circuitry 428 in the downstream PHY shelves 30 .
- Solutions include having the downstream PHY shelf 30 contain a more complex and tighter phase locked loop (PLL) that first derives the MAC clock from timestamps 430 , locks its local PHY clock onto the derived MAC clock, generates a local timestamp, and then locks its local PHY timestamp onto the MAC timestamp. By eliminating the frequency error, the relative timestamp drift would be reduced or eliminated.
- PLL phase locked loop
- timestamp slaves 428 may experience the same relative timestamp drift. This should be acceptable for the upstream PHY shelf 30 as the value of the timestamp is less critical than the downstream PHY shelf 30 and is also not tested at the upstream PHY shelf 32 . This should also be acceptable for the backup MAC 22 _B since it is not typically in use.
- the clocks for the upstream and downstream PHY shelves 30 and 32 can be connected to the external timing shelf 28 as shown in FIG. 28 to achieve clock coherency. If the downstream PHY shelf 30 uses the more complex timestamp synchronizing approach and locks its local clock to the MAC clock, the PHY shelves 30 and 32 might not connect to the external timing shelf 28 .
- the timestamp in the downstream PHY shelf 30 may still be synchronized with the MAC clock in packet shelf 16 .
- the result is that the timestamp in the downstream PHY shelf 30 will get aggressively corrected. For example, it might get corrected every 10 milliseconds (ms) which would be the SYNC interval.
- the cable modem 36 may also update its timestamp accordingly.
- FIG. 32 shows another timing implementation where the downstream PHY shelf 30 operates as the timestamp master 424 and the MAC 22 in the packet shelf 16 operates as a timestamp slave 428 .
- This embodiment may be a simpler and easier solution for the downstream PHY shelf 30 for meeting DOCSIS requirements.
- the downstream PHY shelf 30 allows the downstream PHY shelf 30 to deliver the timestamp 430 to the HFC plant 34 without the relative timestamp drift that could be created in other implementations. It also may simplify clock and timestamp phase locked loop (PLL) circuitry normally needed for a timestamp slave 428 , but not required for the timestamp master 424 . The PLL circuitry may still be required since other downstream PHYs shelves 30 may want to slave off of the timestamp master 424 .
- PLL phase locked loop
- the timestamp of the subsequent downstream PHY shelves 30 may have to be slaved off of the timestamp master 424 in a primary downstream PHY shelf 30 .
- a backup downstream PHY shelf 30 _B may have the same timestamp to prevent the cable modems from rebooting.
- the backup downstream PHY shelf 30 _B may not be in use, so it does not have to meet specifications with respect to timestamp jitter as do other timestamp slaves 428 .
- the backup downstream PHY shelf 30 _B becomes active, it then becomes the timestamp master 424 .
- FIGS. 28-30 use a timing shelf 28 in a hardwired timing configuration that operate as frequency and timestamp masters.
- the system in FIG. 31 implements a timestamp master 424 in the MAC 22 and the system in FIG. 32 implements the timestamp master 424 in the downstream PHY shelf 30 .
- the upstream PHY shelf 32 can operate as the timestamp master 428 .
- the different timestamp master and slave embodiments shown above in FIGS. 28-32 can also be implemented in the cable network shown in FIG. 2 .
- the upstream PHY connectivity may co-exist with the MAC 22 in the same packet shelf 16 .
- the timestamp and frequency signals may be hardwired between the MAC 22 and DOCSIS line card 44 while the timestamp may be sent or received to or from the downstream PHY shelf 30 over packet switched network 26 .
- the clock card in a conventional non-modular CMTS may be reconfigured to operate as the frequency master 422 , frequency slave 426 , timestamp master 424 , or timestamp slave 428 described above in FIGS. 28-32 .
- ranging is used to compensate for the delays that exist between the different shelves 16 , 30 and 32 . Since conventional cable modem ranging is done with respect to measurements taken at the upstream PHY, the amount of correction that will be required should be the difference between the downstream offset from the MAC 22 to the downstream PHY shelf 30 and the upstream offset from the MAC 22 to the upstream PHY shelf 32 . This MAC to PHY ranging is typically performed prior to the conventional cable modem ranging.
- the delay attributed to the modular CMTS architecture is determined by sending a measurement packet 450 A from the MAC 22 to the downstream PHY shelf 30 .
- a measurement packet is shown in FIG. 34 .
- the measurement packet 450 A may contain a timestamp value 454 identifying when it is originally sent by MAC 22 .
- the downstream PHY shelf 30 relays the measurement packet 450 A back to the MAC 22 .
- the difference between an internal received time 460 referenced by the MAC 22 and the original transmit time 454 A determines the roundtrip delay for packet switched network 26 .
- the absolute delay between MAC 22 and downstream PHY shelf 30 may be determined as half of the roundtrip delay for measurement packet 450 A.
- the downstream PHY shelf 30 may add a timestamp value 456 into measurement packet 450 A that indicates when the measurement packet 450 A was initially received over packet switched network 26 and/or another timestamp value 458 indicating when the measurement packet 450 A is output back over packet network 26 to MAC 22 .
- the difference between the timestamp value 456 inserted by downstream PHY shelf 30 and the original timestamp value 454 indicating when the measurement packet 450 A was originally sent by MAC 22 indicates the one-way delay over packet network 26 from MAC 22 to downstream PHY shelf 30 .
- the difference between timestamp value 460 and 458 identifies the one-way delay from the downstream PHY shelf 30 to MAC 22 .
- This absolute delay time from the MAC shelf 22 to the downstream PHY shelf 30 is added to the conventional MAP advance time 414 in FIG. 26 that is normally calculated to account for non-modular CMTS delays related to sending data over the HFC plant 34 . This ensures that MAPs 400 in FIG. 26 arrive at the cable modem 36 in time to be used.
- Another measurement packet 450 B may be sent to determine the transmission delay from the upstream PHY shelf 32 to the MAC 22 .
- the MAC 22 may divide the round trip delay for measurement packet 450 B by half to determine the one-way delay.
- the upstream PHY shelf 32 may add a timestamp value similar to timestamp value 458 indicating when the measurement packet 450 B is sent back to the MAC 22 . Timestamp value 458 is then subtracted from a local MAC receive time, similar to timestamp value 460 in FIG. 34 , indicating when the measurement packet 452 is received by MAC 22 . This difference identifies the one-way delay from upstream PHY shelf 32 to MAC 22 .
- the measurement packets 450 A and 450 B are periodically sent out to provide constant ranging.
- the MAC-PHY delay from packet shelf 16 to downstream PHY shelf 30 and the packet delay from upstream PHY shelf 32 to packet shelf 16 are then dynamically added to the conventional MAP advance time 414 calculated in FIG. 26 . If the delay over packet network 26 varies, the MAP advance time is varied proportionally with the amount of measured network delay. This prevents having to select a worst case network delay.
- the MAP advance value is selected to be at least larger than the MAC-PHY delay time and in one example is calculated as follows: Map Advance Time>[upstream PHY to MAC delay+downstream MAC to PHY delay+conventional MAP advance time 414]
- the timestamp master 424 may be located in one of the downstream PHY shelves 30 .
- the downstream PHY shelf 30 in FIG. 32 may then determine the one-way delay in packet network 26 from the MAC 22 to the downstream PHY shelf 30 and the one-way delay from upstream PHY shelf 30 to MAC 22 . Measurements similar to those described above would be performed by the upstream PHY shelf 32 when the upstream PHY shelf 32 operates as the timestamp master 424 .
- the MAC-PHY delays described above are applicable for the hardwired system described above in FIGS. 28-30 and for the packet switched timestamp systems described above in FIGS. 31 and 32 .
- the hardwired timing system shown in FIG. 28 may not be concerned about timestamp delays.
- the hardwired timing system still has message delays associated with the REQ 136 and GNT 137 messages transported over packet switched network 26 .
- the MAP advance ranging described above in FIGS. 33 and 34 compensate for the delays in sending these DOCSIS messages.
- the timestamp 430 described above in FIGS. 31 and 32 may also be compensated according to the ranging shown in FIGS. 33 and 34 .
- the timestamp value 430 sent in the SYNC message 362 in FIG. 31 may be reduced by the amount of measured delay between MAC 22 and downstream PHY shelf 30 . This allows the timestamp slave 428 to use the same timestamp value originally sent by timestamp master 424 in MAC 22 .
- the delay in packet switched network 26 may exceed DOCSIS requirements which may cause the usable radius of DOCSIS to be significantly decreased. It may then be necessary to take an additional operation that performs the ranging algorithm between the MAC shelf 22 and each of the PHY shelves 30 and 32 . The result of this MAC-PHY ranging then provides PHY shelves 30 and 32 with the same timestamp counter value. The CM ranging then tunes out any inaccuracies left over from the MAC-PHY ranging. This implementation uses a two-way path between the MAC 22 and PHYs 30 and 32 .
- timestamp jitter 471 refers to changes in the amount of time required for timestamps 480 to be transmitted over the packet switched network 26 .
- timestamp packets 480 may be generated by TS master 424 in packet shelf 16 at evenly spaced apart time intervals. However, the packets 480 may arrive at downstream PHY shelf 30 at varying time intervals.
- These varying delays in transmitting and receiving packets 480 over network 26 can be caused by congestion in the output buffer 470 in packet shelf 16 or congestion in the input buffer 483 in the downstream PHY shelf 30 . Jitter is also created by congestion conditions in the switches and routers operating in the packet switched network 26 . This jitter can adversely affect the DOCSIS timing in the modular CMTS system 10 . For example, the timestamp values sent down the HFC plant 34 and used by the cable modems may no longer correspond with the timestamp value originally generated by MAC 22 in packet shelf 16 . This could disrupt the REQ-GNT messaging described above.
- Timestamp packets 480 may initially have some predetermined constant amount of time delay which is accounted for in line 500 . For example, with no packet jitter, the timestamp packets 1 - 6 would be received at times intersecting line 500 . However, due to packet jitter, the packets 1 - 6 can arrive at times that do not intersect line 500 .
- a first timestamp packet # 1 may be received at time 504 A and is longer than the expected receive time corresponding to time 504 B on line 500 .
- the timestamp de-jitter circuit 484 sends the expected timestamp value 504 B to the timestamp rewrite circuit 494 that then replaces the received timestamp value 504 A with timestamp value 504 B.
- the timestamp de-jitter circuit 484 continues to identify the expected timestamp values for the received timestamp values. For example, timestamp # 3 is received at time 508 A before it was expected at time 508 B.
- the de-jitter buffer 484 sends timestamp value 508 B to the timestamp rewrite circuitry 494 that then replaces the timestamp value 508 A with timestamp value 508 B.
- the dejitter circuit 484 continuously tracks the actual received timestamp values. Over time, the received timestamp values may consistently be above or below the dejittered timestamp line 500 . The dejitter circuit 484 may then move line 500 upward or downward according to the actually received timestamp value times. For example, over time timestamps may arrive at times 506 , 510 , etc. that are constantly above the current dejittered timestamp line 500 . The dejitter circuit 484 accordingly adjusts line 500 upward to line 502 . Timestamp line 502 is then used for replacing the received timestamp values with dejittered values. For example, after readjusting to line 502 , a next timestamp # 6 may be received at time 514 A. The de-jitter circuit 484 accordingly replaces the timestamp value 514 A with de-jittered timestamp value 514 B.
- the de-jitter circuit 484 adjusts line 500 downward.
- the new adjusted lower timestamp line is then used for replacing timestamp values.
- the timestamp rewrite circuit 494 replaces the received timestamp values with the timestamp values identified by dejitter circuit 484 .
- the new timestamp values are then added to the DOCSIS data stream 488 or the MPEG data stream from video processing circuit 486 as previously done in conventional CMTS systems.
- the de-jittered timestamp along with the DOCSIS or video data is then sent to QAM 496 , up-converter 498 and then sent over the downstream path of HFC 34 .
- the system described above can use dedicated processor systems, micro controllers, programmable logic devices, or microprocessors that perform some or all of the operations. Some of the operations described above may be implemented in software and other operations may be implemented in hardware.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- General Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
A modular Cable Modem Termination System (CMTS) includes a packet shelf operating a Data Over Cable Service Interface Specifications (DOCSIS) Media Access Control (MAC) framer. One or more downstream Physical Interface (PHY) shelves receive DOCSIS data from the packet shelf over a packet switched network and modulate the DOCSIS data for sending on a downstream path of a cable plant. One or more upstream PHY shelves send DOCSIS data received from an upstream path of the cable plant over the packet switched network to the packet shelf. By separating the PHY components from the MAC and from the system software, the PHY components for a Hybrid Fiber Coax (HFC) plant may be replaced with different PHY components for other access technologies such as wireless, Digital Subscriber Lines (DSL), Ethernet-to-the-Home, Fiber-to-the-Home, or fiber Passive Optical Networks (PONs).
Description
This application is a continuation of U.S. non-provisional patent application Ser. No. 11/134,818 filed May 20, 2005, which claims priority of U.S. provisional patent application No. 60/574,506, filed May 25, 2004, and U.S. provisional patent application No. 60/574,876, filed May 26, 2004, and U.S. provisional patent application No. 60/622,312, filed Oct. 25, 2004, and U.S. provisional patent application No. 60/624,490, filed Nov. 1, 2004, and U.S. provisional patent application No. 60/635,995, filed Dec. 13, 2004, and U.S. provisional patent application No. 60/588,635, filed Jul. 16, 2004, U.S. provisional patent application No. 60/582,732, filed Jun. 22, 2004, and U.S. provisional patent application No. 60/590,509, filed Jul. 23, 2004.
Cable operators have widely deployed high-speed data services on cable television systems. These data services include a cable modem that allows a computer to communicate over an ordinary cable TV network Hybrid Fiber Coax (HFC) cable. A Cable Modem Termination System (CMTS) connects the cable TV network to a data network, such as the Internet. The Data Over Cable Service Interface Specification (DOCSIS) is one of the cable modem standards used for transferring data over the cable TV network.
Increasing demand for cable data services requires additional CMTS processing capacity. This can be prohibitively expensive since each CMTS provides routing, DOCSIS Media Access Control (MAC) processing, downstream signal modulation and upstream signal demodulation. The conventional CMTS architecture does not scale well since any one of the separate components in the CMTS can limit processing capacity and only a limited number of DOCSIS packet processing devices and physical interfaces can be located in the same CMTS chassis.
Different cable networks may also have different processing requirements. For example, one cable network may require substantially more upstream data services than other cable networks. However, it is difficult to customize CMTS architectures for these different data services requirements. It is also expensive to provide redundancy in current CMTS architectures since each backup CMTS includes DOCSIS MAC processors, downstream cable modulators and upstream signal demodulators.
The present invention addresses this and other problems associated with the prior art.
A modular Cable Modem Termination System (CMTS) includes a packet shelf operating a Data Over Cable Service Interface Specifications (DOCSIS) Media Access Control (MAC) framer. One or more downstream Physical Interface (PHY) shelves receive DOCSIS data from the packet shelf over a packet switched network and modulate the DOCSIS data for sending on a downstream path of a cable plant. One or more upstream PHY shelves send DOCSIS data received from an upstream path of the cable plant over the packet switched network to the packet shelf. By separating the PHY components from the MAC and from the system software, the PHY components for a Hybrid Fiber Coax (HFC) plant may be replaced with different PHY components for other access technologies such as wireless, Digital Subscriber Lines (DSL), Ethernet-to-the-Home, Fiber-to-the-Home, or fiber Passive Optical Networks (PONs).
The foregoing and other objects, features and advantages of the invention will become more readily apparent from the following detailed description of a preferred embodiment of the invention which proceeds with reference to the accompanying drawings.
ABBREVIATIONS AND ACRONYMS |
Term | Definition |
CM | Cable Modem |
CMTS | Cable Modem Termination System |
CRC | Cyclic Redundancy Check |
DOCSIS | Data Over Cable Service Interface Specifications |
DS | Downstream |
IP | Internet Protocol |
MAC | Media Access Control. Used to refer to the |
2 element of the system which would include DOCSIS | |
framing and signaling. | |
MPEG | Motion Picture Experts Group |
MPEG-TS | Motion Picture Experts Group Transport Stream |
PHY | Physical Layer. Used to refer to the downstream QAM |
transmitters and the upstream burst demodulators. | |
PHY SHELF | In general, a line card with DOCSIS PHYs and located |
in a switch or router. | |
QAM | Quadrature Amplitude Modulation or Quadrature |
Amplitude Modulator | |
UDP | User Datagram Protocol |
US | Upstream |
The packet shelf 16 includes a Data Over Cable Service Interface Specifications (DOCSIS) packet processor 20 that operates a DOCSIS Media Access Controller (MAC) 22. In this example, a GE interface port 24 is used by the DOCSIS MAC 22 to communicate with the PHY shelf 18 over network 26. A router or routing processor 13 is located either internally or externally with the packet shelf 16. The router 13 transfers packets between the packet shelf 16 and a Wide Area Network (WAN) 12.
The remote PHY 18 includes one or more separate downstream PHY shelves 30, one or more separate upstream PHY shelves 32, and possibly one or more timing shelves 28. It is also possible to locate the timing shelf 28 with the packet shelf 16 or include the timing shelf with the upstream PHY shelf 30 or downstream PHY shelf 32. Any combination of chassis can be used to house the different shelves 28, 30 and 32. In one example, each chassis contains at least one downstream PHY shelf 30 and one upstream PHY shelf 32. In an alternative embodiment, different chassis may contain one or more downstream PHY shelves 30 or one or more upstream PHY shelves 32, but not both. In another embodiment, the remote PHY shelf 18 may contain one or more downstream PHY shelves, while the upstream PHY 32 is combined with the rest of the CMTS in a conventional CMTS chassis.
The PHY shelf 18 is connected to one or more Hybrid Fiber/Coax (HFC) plants 34 that each include a downstream path 40 and an upstream path 42. The HFC 34 is a broadband bidirectional shared-media transmission system that uses fiber trunks between the PHY shelf 18 and fiber nodes (not shown). Coaxial distribution is provided from the fiber nodes to customer locations 38.
The endpoints 38 of the cable network 10 are alternatively referred to as Customer Premise Equipment (CPE) and can include any wired or wireless device that needs to communicate over WAN 12. For example, CPE 38 may include any type of computer server, laptop, Personal Computer (PC), etc that communicates over the HFC 34 through a Cable Modem (CM) 36. The Cable Modem 36 may be located in the CPE 38, may be located in a separate chassis, or may be integrated into a Set Top Box (STB) (not shown). The cable modem 36 operates a DOCSIS MAC that conducts DOCSIS messaging and transfers DOCSIS frames with the DOCSIS MAC 22 in packet shelf 16. Operation of the cable modem 36 is known to those skilled in the art and is therefore not described in further detail.
The modular CMTS 14 decouples the backplane communications that were previously required between a CMTS DOCSIS MAC and the CMTS PHY interface that are used for communicating over cable plant 34. This allows the DOCSIS MAC 22 in packet shelf 16 to communicate to the PHY shelf 18 remotely over packet switched network 26 and allows the downstream PHY shelf 30 to operate independently from the upstream PHY shelf 32. As a result, the modular CMTS components can be more effectively matched with different cable network requirements. For example, if more cable modems 36 are connected to the HFC 34, more downstream PHY shelves 30 and/or upstream PHY shelves 32 can be added to the modular CMTS 14 to support the increased DOCSIS bandwidth demand.
The upstream PHY shelf 32 includes an Ethernet port 32A that both receives DOCSIS messages from the packet shelf 16 and sends DOCSIS messages and data to packet shelf 16. A QAM demodulator 32C demodulates signals received on the upstream path 42 of the HFC 34. A DOCSIS Remote PHY Interface (DRPI) framer 32B frames the data received over upstream path 42 for transport over the packet switched network 26.
In yet another embodiment, the PHY shelves 18_1 and 18_2 also receive video data 50 from one or more video servers 48 over packet switched network 26. In this embodiment, the video server 48 sends video data 50 to the downstream PHY shelves 30_1 and 30_2 for modulation over the downstream path 40 of the different HFCs 34.
This is shown in more detail in FIG. 5 where downstream PHY shelf 30 receives both Moving Picture Experts Group Transport Stream (MPEG-TS) video data 50 from the video server 48 shown in FIG. 4 and also receive the DOCSIS downstream data 54 from the packet shelf 16 (FIG. 4 ). The upstream shelf 32 sends DOCSIS upstream data 56 back to the packet shelf 16 and also generates spectrum management data 58 that may be used by the packet shelf 16, or some other management server, to monitor signaling on the HFC 34. For example, a Digital Signal Processor (DSP) 60 in the upstream shelf 32 may generate analysis data from the frequency spectrum 62 of the upstream path 42. The upstream shelf 32 then converts the analysis data into packets that are then sent over the packet switched network 26 to the packet shelf 16 or other management server.
A remote PHY downstream protocol determines how DOCSIS messages and data is transferred between the MAC 22 in the packet shelf 16 and the downstream PHY shelf 30 and how messaging is sent from the MAC 22 to the upstream PHY shelf 32. In one embodiment, the remote PHY downstream protocol is unidirectional and no acknowledgement messages are required in an upstream direction. In an alternative embodiment, data messages are not acknowledged but control messages are acknowledged.
An upstream scheduler 78 generates DOCSIS MAPs 80, for example, in response to DOCSIS transmit requests 82 received from cable modems 36 (FIG. 1A ). In one example, the MAPs 80 identify timeslots allocated to the requesting cable modem 36 for transmitting data in the upstream path 42 of HFC plant 34 (FIG. 1A ). The MAPs 80 are sent to the MPEG framer 76 and formatted into MPEG frames 86 that are sent to the Ethernet framer 77 along with the MEPG frames 86 containing packet data 70.
It should be understood that the MEPG framer 86 is used in one embodiment of the modular HFC 14. In an alternative embodiment shown in FIG. 6B , a Packet Streaming Protocol (PSP) is used for transporting DOCSIS frames from the packet shelf 16 to the PHY shelf 18. In the PSP embodiment, DOCSIS frames 84 are concatenated together into a continuous data stream 94. The concatenated data stream 94 is then broken up into separate packet fragments 98 that each include a PSP header 96. The PSP headers 96 may include pointers 97 that identify where different DOCSIS frames 84 are located. The concatenation 94 can lower the Packet Per Second (PPS) throughput in switching engines used in the modular CMTS system. The concatenated data stream 94 is then fragmented to meet the Maximum Transfer Unit (MTU) of the packet switched network media 26 (FIG. 1A ).
In yet another embodiment, the DOCSIS frames 84 generated by the DOCSIS framer 74 are not formatted into MPEG frames 86 and alternatively sent directly to the Ethernet framer 77. In the embodiment, the conversion of DOCSIS frames to MPEG frames is performed in the PHY shelf 18.
In one embodiment, the packet shelf 16 creates a tunnel 88 over the packet switched network 26 for transporting the DOCSIS frames to the downstream PHY shelf 30. In one embodiment, an MPEG tunnel 88 is used. In an alternative embodiment, a DOCSIS tunnel is used. If a DOCSIS tunnel is used, a new field 85 may be used in the DOCSIS header for packet fragmentation or concatenation. This allows DOCSIS frames to be fragmented over multiple Ethernet packets. The fragmentation or concatenation field 85 contains numbers identifying the sequence for the fragmented DOCSIS frames 84.
For the MPEG tunnel, the Ethernet framer 77 encapsulates one or more MPEG frames 86 into the payload of the same Ethernet frame 89 that includes an Ethernet header 90 and a Cyclic Redundancy Check (CRC) 92.
An IP connection 102 is established between a device 100 on the WAN 12 and a client or CPE 38 in the cable network 10 through tunnel 88. The tunnel 88 allows the packet shelf 16 to perform the DOCSIS MAC operations and then transport the DOCSIS frames over the packet switched network 26 to the PHY shelf 18. The downstream PHY shelf 30 is therefore not required to perform DOCSIS MAC operations 22 (FIG. 1A ).
UDP Transport Packet Format
In the following protocol usage description, the term “source shelf” refers to the packet shelf 16 and the term “destination shelf” refers to the downstream PHY shelf 30. An Ethernet header 110 is defined by IEEE-802.3. Upon transmission of Ethernet packet 89 by the source shelf, the Ethernet destination address 112 will be the Ethernet address of the destination shelf or of the next hop router in packet switch network 26.
Upon reception of frame 89 by the destination shelf, the Ethernet source address 113 will be the Ethernet address of the output port of the source shelf or of the previous hop router in network 26.
An Ethernet 802.1Q header 114 is defined by IEEE-802.1Q. Header 114 is optional and provides frame prioritization and Virtual Local Area Network (VLAN) support at layer 2 of the Open System Interconnect (OSI) model. In one implementation when header 114 is used, the Length/Type field 115 of the 802.3 header 110 is set to an appropriate 802.1Q TagType value. The packet shelf 16 and the upstream PHY shelf 32 optionally may or may not support header 114.
An IP header 116 is defined by RFC-791. An IP source address 118 in header 116 is the IP address of the source shelf. An IP destination address 120 is the IP address of the destination shelf. The IP destination address 120 may be an IP unicast or IP multicast address. In one example, the IP header 116 uses IPv4 or IPv6. Of course any version of the IP protocol can be used. In one embodiment, the Quality of Service (QoS) information associated with the tunnel 88 is contained in a DiffServ (DS) field 119 of the IP header 116.
A UDP header 122 is defined by RFC-768 and the CRC 132 is defined by IEEE-802.3. Alternatively, a Generic Routing Encapsulation (GRE) header can be used instead of UDP header 122. In one embodiment, the UDP source port 124 is unused and is set by the source shelf to all zeros. The destination shelf may ignore the UDP source port 124. The UDP destination port 126 is a common agreed upon value between the source shelf and the destination shelf. In one implementation, the source shelf might set the UDP checksum 128 to zero and the destination shelf may have the ability to ignore this UDP checksum field.
MPT Mode
The payload of UDP packet 89 for the MPT mode is shown in FIG. 8B . The downstream PHY shelf 30 accepts one to seven (or more for systems with an MTU larger than 1500 bytes) MPEG-TS packets 86 within the UDP payload 130 (FIG. 8A ) of the UDP packet 89. The packet shelf 16 in one embodiment does not put stuffing bytes between the UDP header 122 and the first MPEG-TS header or between consecutive MPEG-TS packets 86.
In one embodiment, the maximum length of 7 MPEG-TS packets 86 are allowed within the UDP/IPv4 packet within a 802.3 Ethernet frame with 802.1Q tagging. Of course other Ethernet frame lengths can also be used and the maximize number of packets allowed can be varied. The downstream PHY shelf 30 may generate NULL MPEG-TS frames when there are no MPEG-TS packets 86 to be transmitted.
PSP Mode
The payload 130 of a UDP packet 89 for PSP mode is shown in FIG. 8C . The PSP Payload Data Unit (PDU) 134C consists of one or more DOCSIS frames. The DOCSIS frame includes a fully formatted DOCSIS header, any extended headers, the DOCSIS payload, and the CRC. The Packet Streaming Protocol (PSP) can take a series of layer 3 packets, layer 2 frames, or any fixed or variable sized pieces of data, assemble them as a stream of back to back Payload Data Units (PDUs) 134D, and then break that stream up into segments and encapsulate each segment with a PSP header 134B. In doing so, the first and last PDU 134D of a PSP packet 89 may be fragmented.
The PSP has two basic modes for recovering individual PDU frames 134 out of a series of PSP packets 134. The two modes are interpretive and non-interpretive. The mode to be used, and the PDU type (for example, DOCSIS, IP, Ethernet), are negotiated as part of the assignment of the payload type.
Interpretive Mode
In the interpretive operation of PSP mode, the PSP packet 134 has one PDU pointer. The PDU pointer points to the first PDU 134D that has its first byte in the PSP packet 134. If the PDU pointer is non-zero, the bytes prior to the location pointed to by the first pointer are considered a fragmented PDU 134D, and are combined with the last PDU of the previous PSP packet 134. The receiver then starts to interpret the contents of the PDU based upon the protocol type associated with the payload type. When it reaches the end of the DOCSIS frame, but it has not reached the end of the PSP packet 134, it assumes another PDU 134D is appended, and begins to interpret that PDU. By doing so, all PDUs 134D are broken apart and reassembled.
Non-Interpretive Mode
In the non-interpretive operation of PSP, the PSP packet 134 has one PDU pointer for each start of a PDU 134D. The same PDU fragment rule applies if the first SOP is non-zero. Since each PDU 134D is explicitly called out, the receiver that manages the reassembly of the PDUs 134D does not have to be aware of the protocol within the PDU. If the last PDU pointer does not have the End bit E set, then the PDU 134D is considered a fragment and is held until the next PSP packet 134 is parsed.
Remote PHY Protocol Operation
MPT Data Path
The Downstream PHY shelf 30 may receive MPEG elementary streams which have been encapsulated in MPT packets 86 (FIG. 8B ) and placed in a UDP datagram. This is referred to as the video MPT flow 30K. A transparent MPT flow 30H is received and transmitted to the QAM interface 30F without any interpretation of the contents of those MPT packets. There may be more than one transparent MPT flow 30H where each flow has a different Differentiated Services Code Point (DSCP) and different addressing, but is destined for the same QAM channel.
All DOCSIS frames, including packet based frames and MAC management based frames, are included within the MPT flow 30J. The downstream PHY shelf 30 searches the MPT payload for any DOCSIS SYNC messages and performs SYNC corrections 30T. It then forwards the MPT packet to the QAM interface 30F. In the MPT mode, MPT frames can be received by the downstream PHY shelf 30 and forwarded directly to the modulation interface 30F without having to terminate and regenerate the MPT framing. Except for manipulation of the payload of the MPT frames for DOCSIS in the SYNC correction 30T.
PSP Data Path
The Packet Streaming Protocol (PSP) is a layer 3 convergence layer protocol which allows packets to be consecutively streamed together and fragmented at arbitrary boundaries. The intent of the PSP mode is to facilitate Quality of Service (QoS). This mode is used for transporting traditional DOCSIS data and signaling messages which use one or more DSCP values. For example, in order to reduce REQ-GNT latency, MAP MAC management messages may be sent using a different DSCP on a different PSP flow 30N than the rest of the DOCSIS channel.
Each PSP flow 30N is received, terminated, and the DOCSIS frames within the flow are extracted by PSP termination 30W. The DOCSIS frames from all the combined PSP flows 30N are sorted into output queues 30P by DSCP Mapping 30V based upon the DSCP value contained within the IP packet in the DOCSIS payload. The outputs of the QoS queues 30P go to a packet scheduler 30Q which decides which queue 30P is to be serviced. The packet scheduler 30Q is also responsible for inserting DOCSIS SYNC messages 30R within the time interval specified by DOCSIS timing 30C. The output of the packet scheduler 30Q goes to a MPT Transmission Convergence (TC) engine 30S that places the DOCSIS frames into MPT frames. The output of MPT engine 30S is sent to the MPT scheduler 30E.
MPT Scheduler
The video MPT flow 30K, the transparent MPT flow 30L, the DOCSIS MPT flow 30M, and the PSP flow 30N provide four MPT flows into the MPT scheduler 30E. The MPT scheduler 30E arbitrates between the four streams and makes the decision which MPT packet will be transmitted at what time. The MPT scheduler 30E receives its scheduling policies from an Edge Resource Management Interface (ERMI) (not shown) and takes into account the DSCP values used within and across the various types of flows.
Addressing
The destination IP address 120 (FIG. 8A ) of the packets 89 is used to choose a downstream PHY shelf 30. The IP address selects a PHY chassis and a PHY chassis may have one or more IP addresses. The destination UDP port 126 is used by the packet shelf 16 and the downstream PHY shelf 30 to select a downstream QAM 30F. More than one destination UDP port 126 may point to the same downstream QAM 30F. Packets destined to different QAMs 30F are addressed to different UDP destination ports 126 if they are on the same IP subnet.
DiffServ Code Point Usage
The DOCSIS frames contain IP packets which have a Differentiated Services Code Point (DSCP). The Type of Service (TOS) bits are a subset of the DSCP bits. The DSCP is a value located in the DiffServ field 119 of the IP header 116 (FIG. 8A ) and is used by routers to determine the correct Per Hop Behaviour (PHB). The DSCP 119 may be used at the egress of the packet shelf 16, within the network 26 from the packet shelf 16 to the downstream PHY shelf 30, and at the ingress of the downstream PHY shelf 30. Examples of traffic that may require flows with different DSCP are signaling, such as DOCSIS MAPs, SYNCs, UCDs; DOCSIS MAC management messages; packet cable signaling messages; data; VoIP; and video.
At the Packet Shelf
For the DOCSIS MPT flow 30M, all signaling and data may have the same DSCP 119. That DSCP 119 may be different than the DSCP used for other network traffic. For transparent MPT flows 30H, each unique flow may have a different DSCP 119. For DOCSIS PSP flows 30N, the PDU DSCP may be defined as the DSCP of the IP Packet contained with the DOCSIS frame contained within the PDU 134D. Different PDU DSCPs may be mapped into different PSP flows 30N where each PSP flow 30N would have a different DSCP 119. More than one PDU DSCP may map to the same PSP flow 30N. Each unique PSP stream 30N is assigned a unique destination UDP port 126.
Network MTU
The packet switched network 26 between the packet shelf 16 and the downstream PHY shelf 30 may have a certain Maximum Transfer Unit (MTU). For example, the MTU for an Ethernet network might be 1522 bytes. One technique for determining this value is to have both endpoints run MTU path discovery. If a maximum size DOCSIS frame were to be tunneled from the packet shelf 16 to the downstream PHY shelf 30, the MTU of the resulting packet would be greater than 1522 bytes. Both the MPT and PSP modes avoid this issue by offering streaming which concatenates and fragments packets.
Early MAP Release
For example, the cable modem 36 sends the data transmit request message 136 through the upstream path 42 of the HFC 34 to the upstream PHY shelf 32. The upstream PHY shelf 32 forwards the request 136 over packet switched network 26 to the DOCSIS MAC 22 in packet shelf 16. The MAC 22 responds with a grant MAP 137 that travels back over the packet switched network 26 to the downstream PHY shelf 30. The downstream PHY shelf 30 forwards the grant MAP 137 over the downstream path 40 of HFC 34 to the cable modem 36. The grant MAP 137 identifies a timeslot in the future allocated to cable modem 36 for transmitting data. The cable modem 36 transmits data 138 over upstream path 42 of HFC 34 during the timeslot allocated in grant MAP 137. The upstream PHY shelf 32 forwards the data 138 to the packet shelf 16 over network 26.
The time required to conduct this ‘request-grant-transmit data’ exchange creates a delay when sending data from the cable modem 36 to another device in the WAN 12 (FIG. 1A). This delay can be further aggravated by the additional delay that may happen transporting the messages 136 and 137 and the data 138 over the packet switched network 26. The delay problem can be even further aggravated by the tunneling scheme described above in FIGS. 6-8 .
Referring to FIG. 9B , the request message 136 is received by the upstream scheduler 78 over the packet switched network 26. The upstream scheduler 78 generates the grant MAP 137 as described above in FIG. 9A .
The MAP 137 is formatted into a DOCSIS frame which is then formatted into an MPEG packet by MPEG framer 76.
The Ethernet framer 77 then combines the MPEG packets containing the MAP 137 with other MPEG packets that may contain other DOCSIS messages, DOCSIS data, or video data into the same payload 130 (FIG. 8A ) of Ethernet packet 89. As described in the example above, Ethernet framer 77 may encapsulate up to seven MPEG packets 138 into the same Ethernet packet 89. In one instance, the Ethernet packets 89 are also rate shaped for transmission at a predetermined rate. This combining of MPEG packets and rate shaping further delays the transmission of MAP 137 to the cable modem 36.
To further minimize the delay, Ethernet packets 144 containing DOCSIS frames 84 (FIG. 6 ) may be given higher priority than Ethernet packets 142 that do not contain DOCSIS frames.
For example, the Ethernet packets 142 may only contain MPEG-TS packets 138 with video data and may accordingly be given lower priority than Ethernet packet 144 that contains DOCSIS MAP message 137.
The packet shelf 16 can contain a high QoS queue 150 and a low QoS queue 152 as described below in FIG. 11 . The MPEG-TS packets 138 containing DOCSIS frames may be loaded into a high QoS queue 150 and the MPEG-TS packets 138 that do not contain DOCSIS frames may be loaded into the low QoS queue 152. The Ethernet packets are output from the high QoS queue 150 with higher priority than the Ethernet packets output from low QoS queue 152.
Quality of Service
Referring to FIG. 11 , packets 154-160 with different Quality of Service (QoS) requirements may be received by the packet shelf 16. For example, video packets 154, data packets 156, voice packets 158 and signaling packets 160 each may have different associated QoS requirements. In one example, the different packets 154-160 each have associated Type Of Service (TOS) values 155 in their IP headers that correspond with the required QoS. The TOS values 155 are alternatively referred to above as Differentiated Services Code Point (DSCP) values.
The MPEG encapsulation described above in FIGS. 6-10 may combine multiple MPEG or DOCSIS packets into the same tunnel. In one implementation, a best effort tunnel 162 is used for transporting the different packets 154-160 over the packet switched network 26. In the best effort implementation, the multiple packets 154-160 are encapsulated into the same Ethernet packet 163 with no consideration of the associated TOS values 155 or other QoS values. Therefore, packets with different TOS values 155 are encapsulated and sent by the packet shelf 16 over the packet switched network 26 in the same tunnel 162.
In a QoS tunnel implementation, the packet shelf 16 includes a processor 20 and multiple packet queues 150 and 152 that are used for storing packets 154-160 according to their corresponding QoS values. In this example, the QoS corresponds to the packet TOS values 155. The queue 150 is associated with high QoS packets and the queue 152 is associated with lower QoS packets. The DOCSIS packet processor 20 receives and loads the different packets 154-160 into either high QoS queue 150 or low QoS queue 152 according to their corresponding TOS values 155.
For example, packets with a TOS value 155 above a predetermined threshold value are loaded into high QoS queue 150 and any packets with a TOS value 155 below the predetermined threshold value are loaded into low QoS queue 152. In this example, the voice packet 154 and several signaling packets 160 have TOS values 155 above the predetermined threshold and are accordingly loaded into high QoS queue 150. The data packets 156 and the video packets 154 have TOS values below the predetermined QoS threshold and are accordingly loaded into low QoS queue 152.
The processor 20 then generates separate high and low QoS tunnels 164 and 166 corresponding to the packets in queues 150 and 152. For example, the voice packet 158 and the signaling packets 160 are encapsulated and transported through the packet switched network 26 over high QoS tunnel 164. The data packets 156 and the video packets 154 in low QoS queue 152 are encapsulated and transported through the network 26 over low QoS tunnel 166.
In another implementation, the number of MPEG or DOCSIS frames that are combined in the different tunnels 164 and 166 may vary according to priority. For example, the high QoS tunnel 164 may encapsulate fewer MPEG or DOCSIS frames together than the lower QoS tunnel 166.
Thus, the packet shelf 16 provides different QoS tunnels to the remote PHYs 18. Any number of different QoS tunnels can be provided according to system requirements. For example, separate tunnels may be established for each different type of packet data 154-160. The processor 20 can also establish the different tunnels according to parameters other than, or in combination with, the TOS value 155. For example, the tunnels 164 and 166 can also be established according to different Packet Identifier (PID) values used for identifying different MPEG streams.
In another embodiment, the different tunnels 164 and 166 may be established according to different User Datagram Protocol (UDP) port values identified in the UDP header 122 (FIG. 8A ). The tunnels may also be established according to any combination of the TOS, PID, and UDP port values. Thus, the tunnels can be established according to any combination of UDP, IP or MPEG parameters or according to the different types of data associated with the packets 154-160.
Latency Management
To prevent packet buildup in queue 176, the packet shelf 16 rate shapes packet traffic to the downstream PHY shelf 30. Packet shelf 16 includes a packet queue 172 that receives packets 170. The downstream PHY shelf 30 includes an input packet queue 174 and the output packet queue 176. The output packet queue 176 has a payload rate of 100% of the QAM bandwidth.
To prevent packets from building up in output queue 176 due to latency, the payload rate of the packet queue 172 in the packet shelf 16 is set to some value less than the 100% of the payload rate for packet queue 176. The packet shelf 16 will then deliver packets at a Variable Bit Rate (VBR) or Constant Bit Rate (CBR) that is less than the maximum payload rate for the output queue 176. This allows the downstream PHY shelf 30 to empty output queue 176 even after a latency period where MPEG nulls have been inserted.
The processor 20 may selectively vary the payload rate capacity of buffer 172 according to an amount of jitter detected in packet switched network 26. For example, a high jitter condition may be detected in packet switched network 26 by the processor 20 using existing DOCSIS ranging operations. The processor 20 then accordingly may reduce the payload rate capacity of packet queue 172. This further reduces the chance of packets backing up in the output queue 176.
In this example, the packets in queues 178 and 180 can in combination have 100% of the payload rate of the output queue 172. Packets sent from high QoS queue 178 are output by the packet shelf 16 with a higher priority than the packets in the low QoS queue 186. The packets from high QoS queue 178 received by the downstream PHY shelf 30 are stored in high QoS queue 184 and the packets from low QoS queue 180 received by the downstream PHY shelf 30 are stored in low QoS queue 186. The packets in low QoS queue 186 can be delayed or even dropped when packets backup in output queue 176 due to latency conditions.
The packets 203 sent from the upstream PHY shelf 32 to the packet shelf 16 containing the DMPI blocks 200 are referred to generally as DMPI over IP (DoIP) packets 203. The upstream PHY protocol used for transporting the DoIP packets 203 includes both control plane and forwarding plane modes. In another embodiment, the upstream PHY shelf receives DMPI blocks from the PHY, assembles them into DOCSIS packets, and forwards those DOCSIS packets to the M-CMTS using the same packet format that is used in the downstream direction.
Forwarding Plane In the forwarding plane, the upstream PHY shelf 32 takes content off the HFC plant 34 and sends it over the network 26 to the packet shelf 16. Cable modems 36 (FIG. 1A ) send upstream DOCSIS packets 202 in the form of Forward Error Correction (FEC) blocks mapped to a QAM payload that are then demodulated by the QAM receivers in PHYs 32C and converted into DMPI blocks 200.
The DOCSIS REMOTE PHY Interface (DRPI) framer 32B encapsulates the DMPI blocks 200 into Ethernet frames as described above in FIG. 8A and then sends the DoIP packets 203 over the packet switched network 26 to packet shelf 16. As described above, in one embodiment, the packet switched network 26 is an Ethernet network. However, any IP based network may be used. The upstream PHY shelf 32 may encapsulate multiple DMPI blocks 200 received on the same upstream paths 42 into the same DoIP packets 203 or may alternatively encapsulate DMPI blocks 200 generated from different upstream paths 42A-42C into the same DoIP packets 203.
A single DoIP packet 203 (FIG. 14 ) may contain DMPI blocks 200 from more than one upstream path 42 and more than one logical channel on the same upstream path 42. The UDP destination port 126 (FIG. 8A ) represents a group of one or more physical upstream channels. The upstream PHY shelf 32 in one embodiment does not put stuffing bytes between the UDP header 112 (FIG. 8A ) and the header in the first DMPI block 200 or between consecutive headers of the DMPI blocks 200.
DMPI Data Blocks
The format of the DMPI blocks 200 is defined in the DOCSIS 2.0 specification and is therefore not described in further detail. The supported block types are FIRST_DATA block, MIDDLE_DATA block, LAST_DATA block, PHY_STATUS block, and NO_BURST block. In one embodiment, the CHANNEL block is discarded and a DOMAIN block 208 is included as described below in FIG. 16 .
Referring to FIG. 16 , the DMPI domain block 208 is a new block definition and may not be supported by a conventional native DMPI interface on the upstream PHY chip. The conventional DMPI specification defined in DOCSIS 2.0 is per channel. The DMPI domain block 208 allows DMPI blocks 200 from multiple channels 42A-42C (FIG. 14 ) to be multiplexed together into one DoIP packet 203, and provides a way for the packet shelf 16 to sort incoming intermixed DMPI blocks 200 into correct upstream channel queues.
The DMPI domain block 208 is inserted once at the beginning of a Type-Length-Value (TLV) section of the UDP payload 204 (FIG. 15 ). The DMPI domain block 208 is inserted for example when a next DMPI data block 200 is from a different logical or physical upstream channel 42 than the previous DMPI block 200. The DMPI domain block 208 may contain a domain ID that identifies the particular PHY interface 32C and HFC 34 associated with the DMPI block 200 and an Upstream Channel ID (UCID) corresponds to the unique logical channel ID that the DOCSIS data 202 is received over. Thus, every unique logical upstream channel connected to every PHY 32C (FIG. 14 ) is uniquely identified.
Even though the packet shelf 16 publishes the timestamp snapshot 210, it is created by the upstream PHY shelf 32 by the PHY 32C. The upstream PHY shelf 32 then periodically sends the snapshot 210 to the packet shelf 16.
In the modular CMTS 14, the packet shelf 16 maintains a copy of the frame counter 212 and mini-slot counter 214 for each logical upstream channel which is time aligned with its copy of the timestamp counter 216. These three counters are used to create the timestamp snapshot 210 in an Upstream Channel Descriptor (UCD) message 364 (FIG. 25 ). The packet shelf can alternatively use the snapshot directly from the PHY shelf. The upstream PHY shelf 32 then either periodically sends a copy of its timestamp snapshot 210 to the packet shelf 16 or sends a copy during different conditions, such as during initialization of the upstream channel or when the UCD changes.
The packet shelf 16 time aligns the frame counter 212 and the mini-slot counter 214 from the upstream PHY self timestamp snapshot 210 to the packet shelf timestamp counter and stores those values as the current frame counter and mini-slot counter values.
Packet Drop and Misorder Recovery
An “open” DMPI data block 200 sequence is when a first block has been received but a last block has not been received (if PHY_STATUS is not expected) or if an expected PHY_STATUS has not been received. In the absence of dropped DoIP packets 203, an open sequence may be closed by reception of the LAST_DATA block and the optional PHY_STATUS block. For example, if the PHY_STATUS block is being used, a block sequence is not completely closed until the PHY_STATUS block is received by the packet shelf 16.
If the packet shelf 16 receives out-of-sequence DoIP packets 203, it waits a certain amount of time to determine if the packet or packets have been misordered (i.e. the missing packets are received eventually) or lost (they are not received within the timeout period). If the packet shelf 16 determines that one or more DoIP packets 203 have been lost, the DMPI blocks 200 from all open block sequences are dropped until these open sequences can be terminated. Termination occurs by the reception of new FIRST_DATA block from the same PHYs 32C and logical channels as the block sequences affected by the dropped packets 203.
Control Plane
Quality of Service
Referring to FIGS. 14 and 18 , DOCSIS packets 202 are received by the upstream PHY shelf 32. The DOCSIS packets 202 include a DOCSIS header 224, an extended header field 226, Ethernet header 228, IP header 218, an application dependant UDP header 230, a tunnel header such as RTP or L2TPv3 232, payload 234, and CRC 236. The DOCSIS packets 202 can have different QoS priorities. For example, different QoS values 222 may be contained within the IP header 218 received by the upstream PHY shelf 32.
However, a portion 220 of the DOCSIS packet 202 may be encrypted by the cable modem 36. Alternatively, some of the header information that includes the QoS priority information 222 may be suppressed in the DOCSIS packet 202. In either situation, the upstream PHY shelf 32 may not be able to identify the QoS priority information 222. The challenge then is how to maintain priority for the DOCSIS packets 202 when they are sent over the packet switched network 26 to the packet shelf 16.
During a provisioned QoS to SID assignment, software in the packet shelf 16 sends the SID-TOS mappings 254 to the upstream PHY shelf 32 for all the different SID values 253 that may be used by PHY 32C (FIG. 14 ). For example, the SID-TOS mappings 254 may be sent during initialization, registration, or dynamic service setup, and stored in table 252. The upstream PHY shelf 32 then uses the table 252 to assign the DOCSIS packets 202 priorities as described below. In one ToS overwrite feature, the CMTS forces a ToS corresponds to a DOCSIS priority on the packets sent from the cable modem.
In an alternative dynamic QoS to SID assignment, the packet shelf 16 dynamically sends different SID-TOS values 255 to the upstream PHY shelf 32 in DRPI MAPs 256. FIG. 20 shows the DRPI MAP 256 in more detail. The MAP 256 is conventionally used for identifying mini-slots for cable modem upstream transmissions and includes a conventional DRPI header 258, SID value 260, mini-slot value 263, and mini-slot length value 264.
However, the MAP 256 now also includes new SID to TOS mapping 266 that dynamically associates a TOS value to the SID values 253 for DOCSIS packets received on the HFC 34. The PHY 32C in the upstream PHY shelf 32 uses the SID value 260 and mini-slot values 262 to demodulate data bursts from cable modems. The DRPI framer 32B then uses the SID-TOS mapping 266 associated with the demodulated DMPI blocks 200 (FIG. 14 ) to prioritize the DoIP packets 203 sent over network 26 (FIG. 19 ).
The SID-TOS MAPs 256 allows the DRPI framer 32B (FIG. 14 ) to only keep track of a relatively small number of SID-TOS values 266. Further, if a TOS value 255 for a particular SID 253 is dynamically changed during transmission of MAP 256, there are no synchronization problems that arise with other upstream PHY shelves 32. There may be no need to send a new mapping each time a modem is registered or a dynamic service is set up.
For example, the DMPI blocks 200 associated with SID values 1 and 4 have high TOS values and are accordingly loaded by the DRPI framer 32B into the high QoS queue 272. The DRPI framer 32B encapsulates the DMPI blocks 200 from high QoS queue 272 into a DoIP packet 278 assigned a high QoS value. For example, a high TOS or DSCP value 282 is assigned to the DS field 119 in the IP header 116 (FIG. 8A ) of DoIP packet 278.
The DMPI blocks 200 having SID values 270 associated with low TOS values 255 are loaded by PHY 32C into the low QoS queue 276. The DRPI framer 32B then encapsulates the DMPI blocks 200 in low QoS queue 276 into a low QoS DoIP packet 280 having a low TOS or DSCP value 282.
In an alternative embodiment shown in FIG. 22 , the upstream PHY shelf 32 uses a best effort mode where DMPI blocks 200 are encapsulated into DoIP packets 292 without any consideration for associated TOS values. However, request messages may still be sent in separate high priority DoIP packets as described below in FIG. 23 .
Early Request Extraction
Referring to FIG. 23 , DOCSIS frames 202 can be sent by the cable modems 36 (FIG. 1A ) in many different formats and can contain different types of information. For example, the DOCSIS frames can contain data 308 or bandwidth request (REQ) messages 311. As described above in FIG. 9A , the cable modems 36 send REQ messages 311 to the packet shelf 16 to request data transmission in the upstream path 42 of the HFC 34.
This transmission REQ message 311 may be sent in a separate DOCSIS REQ frame 310 or may be piggybacked along with data in DOCSIS frame 312. In another embodiment, the REQ message 311 is sent along with multiple concatenated data packets in DOCSIS frame 314. It is also possible that the REQ message 311 may be embedded in one of the data concatenated data packets in DOCSIS frame 314. In yet another embodiment, the REQ message 311 is combined with fragmented DOCSIS frames 316A-316C.
The PHY 32C converts the DOCSIS frames 308-316 into DMPI blocks 200. As described above, the DRPI framer 32B then encapsulates multiple DMPI blocks 200 together into DoIP packets that are then transported over the network 26. As also previously described above in FIGS. 9A and 9B , it is desirable to reduce the delay from when a cable modem 36 sends a REQ message 311 to the time a grant MAP is received back from the packet shelf 16.
However, the encapsulation of multiple DMPI blocks 200 into DoIP packets or tunnels can delay this REQ-GNT-Transmit process. For example, a REQ message 311 may be the first DMPI block 200 encapsulated in a DoIP packet that contains multiple DMPI blocks 200. The REQ message 311 could be therefore be delayed until several other DOCSIS frames 202 are received, converted into DMPI blocks 200, and then encapsulated into the same DoIP packet.
To reduce the REQ-GNT-Transmit delay time, the upstream PHY shelf 32 conducts an early REQ extraction. In one embodiment, the PHY 32C monitors the bytes of the incoming DOCSIS frames 202 for REQ messages 311. Whenever a REQ message 311 is detected, the PHY 32C generates a separate DMPI REQ block 318 that is sent to a separate REQ message queue 306. As soon as the DMPI REQ block 318 is received in REQ queue 306, the DRPI framer 32B formats the REQ block 318 into a DoIP packet 320 and sends it over the network 26 to the packet shelf 16. This eliminates the possible delay that could be created encapsulating the DMPI REQ block 318 with other DMPI blocks 200.
To further reduce the REQ-GNT-Transmit delay, the DoIP REQ packet 320 is assigned a high priority QoS or DSCP value. For example, the DRPI framer 32B assigns a high TOS value 322 to DoIP REQ packet 320. This allows packet 320 to be processed with higher priority through the packet switched network 26 (FIG. 1A ).
The REQ message 311 may be replicated in DMPI REQ block 318 and the original REQ message 311 processed in a normal manner. In other words, the REQ message 311 is also converted into DMPI blocks 200, loaded into a data queue 304 in the DRPI framer 32B along with other DOCSIS data, and then encapsulated with other DOCSIS data 326 into DoIP packet 324. The DoIP packet 324 may or may not be assigned a particular TOS value 328 based on the priority criteria discussed above in FIGS. 21 and 22 . Thus, the same REQ message 311 may be sent in both the DoIP REQ packet 320 and the DoIP packet 324.
It is likely, but not guaranteed, that the DoIP REQ packet 320 will be received by the packet shelf 16 before DoIP packet 324. In either case, the packet shelf 16 may process whichever REQ message 311 is received first in packet 320 or 324.
It is possible that the PHY 32C may not be able to extract and separately transmit a DoIP REQ packet 320 for all received REQ messages 311. In this situation, the PHY 32C may send DRPI framer 32B an indication whenever a REQ message 311 is successfully detected. The DRPI framer may then insert an indicator in a designated field in either or both of the DoIP REQ packet 320 and DoIP packet 324 that notifies the packet shelf 16 that the same REQ message 311 has been sent in two different DoIP packets. The packet shelf 16 then ignores the second received REQ message 311.
The upstream PHY shelf 32 can also use the configuration in FIG. 24 in combination with another embodiment of the early REQ release. In this embodiment, the REQ messages 311 are encapsulated along with other DMPI blocks 200A-200C by the DRPI framer 32B into the same DoIP packet 346. However, the DoIP packet 346 is released as soon as the DMPI block 200 containing the REQ message 311 is encapsulated. The next DoIP packet 348 then starts encapsulating DMPI blocks 200 where DoIP packet 346 left off.
The DoIP packet 346 containing the REQ message 311 may optionally be tagged with a high TOS value 350 by the DRPI framer 32B. If other REQ messages 311 are detected from other upstream paths 42A-42C, they may be combined with the REQ message 311 already loaded into DoIP packet 346.
The DOCSIS MAC 20 in the packet shelf 16 programs the downstream PHY shelf 30 and upstream PHY shelf 32 by sending MAP messages 360, SYNC messages 362 and Upstream Channel Descriptor (UCD) messages 364. The MAP messages 360 are used for allocating timeslots as described above. The SYNC messages 362 are used for timing and the UCD messages 364 contain programming parameters for the PHY shelves 30 and 32. The messages 360, 362 and 364 can be encapsulated into tunnels as described above. The specific contents of the MAP, SYNC, and UCD messages are described in the DOCSIS 2.0 specification and are therefore not described in further detail.
In another embodiment, the contents of the MAP, UCD, and SYNC messages can be sent with the control messages of a tunneling protocol. For example, the MAP, UCD, and SYNC messages could be represented in data fields in Attribute Value Pairs (AVPs) with the Layer 2 Tunneling Protocol version 3 (L2TPv3).
Referring back to FIG. 9A , with the separation of the PHY shelves 30 and 32 from the MAC 22, there is the potential penalty of increased delay from when a REQ 136 is issued by the cable modem 36, and when the cable modem 36 eventually receives the GNT message 137. In this example, the REQ message 136 can represent either a DOCSIS request Information Element (IE) that is successfully sent through a contention request slot, or a piggyback request. The GNT 137 is used for the explicit scheduling of an upstream packet and is typically sent in the DOCSIS MAP message as shown in FIG. 9B .
Once the REQ 406 with delay 408 is received by a non-modular CMTS 398, it goes through the physical interface (PHY), an input queuing process, and finally is processed by an upstream scheduler 78 shown in FIG. 6 . The upstream scheduler 78 may decide to delay the packet for QoS reasons such as rate shaping or because it has decided that other REQs have higher priority at that moment.
Assuming that this delay either does not happen or is over, the upstream scheduler 78 creates a MAP 400 for placing a GNT 412. To do this, the upstream scheduler 78 creates a MAP message 400B that represents some time in the future. The time and associated MAP message 400B is chosen using a multi-step process. First, a parameter called the MAP advance time 414 is added to a current time reference (timestamp). Then, a further addition of time is added to get to the next MAP boundary 400B. Then the packet data Information Element (IE) is scheduled within that MAP boundary 400B. The cycle then continues.
The MAP advance time 414 is used by the CMTS 398 to allow for a variety of time delays in the round trip path both internal and external between the CMTS 398 and the cable modem 399. One of the values used in the MAP advance time 414 is derived either statically or dynamically and takes into account the delay required to send signals over the HFC plant 34 (FIG. 1A ) and provides a design margin referred to as the “safety” factor.
In the conventional non-modular CMTS 398, the MAP advance time 414 takes into account the HFC plant round trip time, cable modem 399 minimum processing delay, CMTS 398 receive queue delay, CMTS 398 transmit queue delay, CMTS 398 transmit PHY delay (interleaver), and some margin.
However, the modified REQ-GNT data path for the remote PHY system 10 shown in FIGS. 1A and 9A has additional delays associated with the processing time of the downstream remote PHY 30, upstream remote PHY 32, MAC 22, and latency in the packet switched network 26. The timing and synchronization system described below provides DOCSIS timing and synchronization for the remote PHY architecture 10 and compensates for the additional delays created in the modular CMTS.
Timing Shelves
If the packet shelf 16 and PHY shelves 30 and 32 happen to be in the same physical location, then the same timing shelf 28 might be used for all the different CMTS shelves. If any combination of the packet shelf 16, downstream PHY shelf 30, or upstream PHY shelf 32 are located remotely from each other, then separate timing shelves 28 may be used at each remote location. Global positioning system (GPS) receivers 430 are connected to each timing shelf 430 and provide a frequency reference and clock used for synchronizing the different autonomously operating timing shelves 28.
Sending a Software Timestamp Over a Packet Switched Network
Sending a DOCSIS timestamp is one embodiment. Sending a timestamp value that is not in DOCSIS format is also possible. The main distinction is that the timestamp is sent in a software message over a network rather than over a dedicated timing interface.
The timing system in FIG. 31 separates frequency and timestamp generation. It is assumed that the packet shelves 16, and PHY shelves 30 and 32, can run stand-alone. In this timing configuration, there may be hardwired point to point connections between different frequency circuits 422, but the timestamp circuits 424 and 428 typically communicate via the packet switched network 26.
In one embodiment, a DOCSIS timestamp is sent. In an alternative embodiment, a timestamp value is sent that is not in the DOCSIS format. The distinction is that the timestamp is sent in a software message over a network rather than over a dedicated timing interface.
For advanced time division multiple access (ATDMA/TDMA), each shelf 16, 30 and 32 may generate its own internal clock frequency using a frequency master 422. These internal frequency references each drive an internal timestamp counter. For synchronous code division multiple access (SCDMA) systems, a timing shelf 28 operating a frequency master module 422 may supply a hardwired clock signal 423 to frequency slaves 426 similar to that shown in FIG. 28 . Alternatively, a frequency module in one of the shelves 16, 30 or 32 may operate as a frequency master 422 and then supply a hardwired clock signal 423 to the other packet shelves 16 or PHY shelves 30 and 32 in the same physical location.
The packet switched network 26 eliminates some of the physical limitations of conventional CMTS systems. In conventional CMTS systems timing and clock signals have to be either star-wired or daisy chained between multiple CMTS shelves in the same chassis. The star-wired connectively requires separate ports for each slave component and the daisy chained connectivity can cause the clock to degrade when passing through each CMTS component. However, the packet switched connectivity shown in FIG. 31 only requires a single network connection in the timestamp master 424 and timestamp slaves 428 for exchanging timestamp information.
The internal frequencies of each shelf 16, 30 and 32 may be slightly different and the timestamp values for each system component may drift from each other. To address this problem, one of the timestamp generators operates as the timestamp master 424. In this example, the timestamp generator in the packet shelf 16 is the timestamp master 424 and the other timestamp counters in the downstream PHY shelf 30 and the upstream PHY shelf 32 operate as timestamp slaves 428. The timestamp slaves 428 continuously receive timestamps 430 over the IP network 26 from the timestamp master 424, dejitter them, and then update their timestamp counters accordingly. In other embodiments, one of the downstream PHY shelves 30 or upstream PHY shelves 32 may operate as the timestamp master 424.
Since the data path between the MAC card 22 and the downstream PHY shelf 30 may contain jitter, the downstream PHY shelf 30 overwrites the timestamp value 430 in the DOCSIS stream with a corrected and de-jittered timestamp. This is referred to as DOCSIS timestamp correction and is described in more detail below in FIGS. 35 and 36 .
In the example of FIG. 31 , a primary MAC 22 operates as the timestamp master 424 and a backup MAC 22_B operates as a timestamp slave 428. If the primary operating MAC 22 fails, the backup MAC 22_B is brought on line without disrupting the timestamp value. For example, whenever the timestamp slave 428 in MAC 22_B does not receive the timestamp 430 after some period of time, it may automatically convert over to operating as the timestamp master 424. The backup timestamp master operation provided by MAC 22_B can alternatively be implemented in one of the PHY shelves 30 or 32.
In one embodiment, the timestamp 430 is sent by MAC 22 to the PHY shelves 30 and 32 in a DOCSIS SYNC message 362 (see FIG. 25 ). The SYNC message 362 can be embedded within the data stream 88 that runs between the MAC card 22 and the downstream and upstream PHY shelves 30 and 32 as shown in FIGS. 6-8 . The data stream 88 may use the same packet format used in video on demand (VOD) systems for sending content to edge QAMs.
A UDP packet 89 shown in FIG. 8A may be used for carrying the timestamp 430 to the PHY shelves 30 and 32. As described in FIG. 6 above, the UDP packet 89 may contain multiple MPEG-TS frames 86. To minimize the jitter, the MAC 22 may place the SYNC message 362 in the same position each time within the data stream 88. To minimize delay between the MAC 22 and the downstream PHY shelf 30, the MAC 22 may place the SYNC message 362 in a last MPEG-TS frame 86 in the UDP packet 89 shown in FIG. 8A .
The SYNC message 362 may be sent over the packet switched network 26 in separate unicast packets having IP addresses associated with the different packet or PHY shelves 30 and 32. Alternatively, the MAC 22 may send a single SYNC message 362 in a multicast packet that is received by each of the packet and PHY shelves 30 and 32.
Relative Timestamp Drift
Still referring to FIG. 31 , the timestamp slaves circuits 428 in the upstream PHY shelf 32 and in the backup MAC 16_B may be simpler than the timestamp slave circuitry 428 in the downstream PHY shelves 30. This is due to the downstream PHY shelf 30 having to meet the DOCSIS specification of less than 500 nanoseconds (ns) of timestamp jitter. Since the timestamp in the remote PHY shelves 30 and 32 are synchronized with the timestamp 430 in the MAC 22, which in turn is derived from a MAC clock, the downstream PHY timestamp will appear to drift with respect to its local downstream PHY clock.
Solutions include having the downstream PHY shelf 30 contain a more complex and tighter phase locked loop (PLL) that first derives the MAC clock from timestamps 430, locks its local PHY clock onto the derived MAC clock, generates a local timestamp, and then locks its local PHY timestamp onto the MAC timestamp. By eliminating the frequency error, the relative timestamp drift would be reduced or eliminated.
It should be noted that all timestamp slaves 428, including the timestamp slave 428 in the upstream PHY shelf 32 and the backup MAC 22_B, may experience the same relative timestamp drift. This should be acceptable for the upstream PHY shelf 30 as the value of the timestamp is less critical than the downstream PHY shelf 30 and is also not tested at the upstream PHY shelf 32. This should also be acceptable for the backup MAC 22_B since it is not typically in use.
If the downstream PHY shelf 30 uses the simpler timestamp synchronizing approach and does not lock its local clock to the MAC clock, the clocks for the upstream and downstream PHY shelves 30 and 32 can be connected to the external timing shelf 28 as shown in FIG. 28 to achieve clock coherency. If the downstream PHY shelf 30 uses the more complex timestamp synchronizing approach and locks its local clock to the MAC clock, the PHY shelves 30 and 32 might not connect to the external timing shelf 28.
Even if the upstream and downstream PHY shelves 30 and 32 are on the same clock, the timestamp in the downstream PHY shelf 30 may still be synchronized with the MAC clock in packet shelf 16. The result is that the timestamp in the downstream PHY shelf 30 will get aggressively corrected. For example, it might get corrected every 10 milliseconds (ms) which would be the SYNC interval. The cable modem 36 may also update its timestamp accordingly.
Timestamp Master in PHY Shelf
Implementing the timestamp master 424 in the downstream PHY shelf 30, allows the downstream PHY shelf 30 to deliver the timestamp 430 to the HFC plant 34 without the relative timestamp drift that could be created in other implementations. It also may simplify clock and timestamp phase locked loop (PLL) circuitry normally needed for a timestamp slave 428, but not required for the timestamp master 424. The PLL circuitry may still be required since other downstream PHYs shelves 30 may want to slave off of the timestamp master 424.
For example, there may be multiple downstreams per MAC domain where those downstreams feed PHYs shelves in different chassis, and where it was required to load balance across those downstreams without re-ranging. The timestamp of the subsequent downstream PHY shelves 30 may have to be slaved off of the timestamp master 424 in a primary downstream PHY shelf 30.
There also may be multiple downstream and upstream PHYs shelves 30 and 32, respectively, per MAC domain, where PHY shelves are swapped either with a RF switch or through load balancing. In this situation, a backup downstream PHY shelf 30_B may have the same timestamp to prevent the cable modems from rebooting. In the case of the RF switch, the backup downstream PHY shelf 30_B may not be in use, so it does not have to meet specifications with respect to timestamp jitter as do other timestamp slaves 428. When the backup downstream PHY shelf 30_B becomes active, it then becomes the timestamp master 424.
In summary, the systems described in FIGS. 28-30 use a timing shelf 28 in a hardwired timing configuration that operate as frequency and timestamp masters. The system in FIG. 31 implements a timestamp master 424 in the MAC 22 and the system in FIG. 32 implements the timestamp master 424 in the downstream PHY shelf 30. In yet another embodiment, the upstream PHY shelf 32 can operate as the timestamp master 428.
The different timestamp master and slave embodiments shown above in FIGS. 28-32 can also be implemented in the cable network shown in FIG. 2 . In this configuration, the upstream PHY connectivity may co-exist with the MAC 22 in the same packet shelf 16. In this embodiment, the timestamp and frequency signals may be hardwired between the MAC 22 and DOCSIS line card 44 while the timestamp may be sent or received to or from the downstream PHY shelf 30 over packet switched network 26.
In a retrofit situation, the clock card in a conventional non-modular CMTS may be reconfigured to operate as the frequency master 422, frequency slave 426, timestamp master 424, or timestamp slave 428 described above in FIGS. 28-32 .
MAC-PHY Ranging
Referring to FIG. 33 , ranging is used to compensate for the delays that exist between the different shelves 16, 30 and 32. Since conventional cable modem ranging is done with respect to measurements taken at the upstream PHY, the amount of correction that will be required should be the difference between the downstream offset from the MAC 22 to the downstream PHY shelf 30 and the upstream offset from the MAC 22 to the upstream PHY shelf 32. This MAC to PHY ranging is typically performed prior to the conventional cable modem ranging.
In one embodiment, the delay attributed to the modular CMTS architecture is determined by sending a measurement packet 450A from the MAC 22 to the downstream PHY shelf 30. One example of a measurement packet is shown in FIG. 34 . The measurement packet 450A may contain a timestamp value 454 identifying when it is originally sent by MAC 22. The downstream PHY shelf 30 relays the measurement packet 450A back to the MAC 22.
The difference between an internal received time 460 referenced by the MAC 22 and the original transmit time 454A determines the roundtrip delay for packet switched network 26. The absolute delay between MAC 22 and downstream PHY shelf 30 may be determined as half of the roundtrip delay for measurement packet 450A.
Alternatively, the downstream PHY shelf 30 may add a timestamp value 456 into measurement packet 450A that indicates when the measurement packet 450A was initially received over packet switched network 26 and/or another timestamp value 458 indicating when the measurement packet 450A is output back over packet network 26 to MAC 22.
The difference between the timestamp value 456 inserted by downstream PHY shelf 30 and the original timestamp value 454 indicating when the measurement packet 450A was originally sent by MAC 22 indicates the one-way delay over packet network 26 from MAC 22 to downstream PHY shelf 30. The difference between timestamp value 460 and 458 identifies the one-way delay from the downstream PHY shelf 30 to MAC 22.
This absolute delay time from the MAC shelf 22 to the downstream PHY shelf 30 is added to the conventional MAP advance time 414 in FIG. 26 that is normally calculated to account for non-modular CMTS delays related to sending data over the HFC plant 34. This ensures that MAPs 400 in FIG. 26 arrive at the cable modem 36 in time to be used.
Another measurement packet 450B may be sent to determine the transmission delay from the upstream PHY shelf 32 to the MAC 22. The MAC 22 may divide the round trip delay for measurement packet 450B by half to determine the one-way delay. Alternatively, the upstream PHY shelf 32 may add a timestamp value similar to timestamp value 458 indicating when the measurement packet 450B is sent back to the MAC 22. Timestamp value 458 is then subtracted from a local MAC receive time, similar to timestamp value 460 in FIG. 34 , indicating when the measurement packet 452 is received by MAC 22. This difference identifies the one-way delay from upstream PHY shelf 32 to MAC 22.
The measurement packets 450A and 450B are periodically sent out to provide constant ranging. The MAC-PHY delay from packet shelf 16 to downstream PHY shelf 30 and the packet delay from upstream PHY shelf 32 to packet shelf 16 are then dynamically added to the conventional MAP advance time 414 calculated in FIG. 26 . If the delay over packet network 26 varies, the MAP advance time is varied proportionally with the amount of measured network delay. This prevents having to select a worst case network delay.
The MAP advance value is selected to be at least larger than the MAC-PHY delay time and in one example is calculated as follows:
Map Advance Time>[upstream PHY to MAC delay+downstream MAC to PHY delay+conventional MAP advance time 414]
Map Advance Time>[upstream PHY to MAC delay+downstream MAC to PHY delay+conventional MAP advance time 414]
As shown above in FIG. 32 , the timestamp master 424 may be located in one of the downstream PHY shelves 30. The downstream PHY shelf 30 in FIG. 32 may then determine the one-way delay in packet network 26 from the MAC 22 to the downstream PHY shelf 30 and the one-way delay from upstream PHY shelf 30 to MAC 22. Measurements similar to those described above would be performed by the upstream PHY shelf 32 when the upstream PHY shelf 32 operates as the timestamp master 424.
The MAC-PHY delays described above are applicable for the hardwired system described above in FIGS. 28-30 and for the packet switched timestamp systems described above in FIGS. 31 and 32 . For example, the hardwired timing system shown in FIG. 28 may not be concerned about timestamp delays. However, referring back to FIG. 9A , the hardwired timing system still has message delays associated with the REQ 136 and GNT 137 messages transported over packet switched network 26. The MAP advance ranging described above in FIGS. 33 and 34 compensate for the delays in sending these DOCSIS messages.
The timestamp 430 described above in FIGS. 31 and 32 may also be compensated according to the ranging shown in FIGS. 33 and 34 . For example, the timestamp value 430 sent in the SYNC message 362 in FIG. 31 may be reduced by the amount of measured delay between MAC 22 and downstream PHY shelf 30. This allows the timestamp slave 428 to use the same timestamp value originally sent by timestamp master 424 in MAC 22.
For example, the delay in packet switched network 26 may exceed DOCSIS requirements which may cause the usable radius of DOCSIS to be significantly decreased. It may then be necessary to take an additional operation that performs the ranging algorithm between the MAC shelf 22 and each of the PHY shelves 30 and 32. The result of this MAC-PHY ranging then provides PHY shelves 30 and 32 with the same timestamp counter value. The CM ranging then tunes out any inaccuracies left over from the MAC-PHY ranging. This implementation uses a two-way path between the MAC 22 and PHYs 30 and 32.
Timestamp Correction
Referring to FIG. 35 , timestamp jitter 471 refers to changes in the amount of time required for timestamps 480 to be transmitted over the packet switched network 26. For example, timestamp packets 480 may be generated by TS master 424 in packet shelf 16 at evenly spaced apart time intervals. However, the packets 480 may arrive at downstream PHY shelf 30 at varying time intervals.
These varying delays in transmitting and receiving packets 480 over network 26 can be caused by congestion in the output buffer 470 in packet shelf 16 or congestion in the input buffer 483 in the downstream PHY shelf 30. Jitter is also created by congestion conditions in the switches and routers operating in the packet switched network 26. This jitter can adversely affect the DOCSIS timing in the modular CMTS system 10. For example, the timestamp values sent down the HFC plant 34 and used by the cable modems may no longer correspond with the timestamp value originally generated by MAC 22 in packet shelf 16. This could disrupt the REQ-GNT messaging described above.
The downstream PHY shelf 30 uses a timestamp de-jitter circuit 484 and a timestamp rewrite circuit 494 to compensate for the packet jitter 471. FIG. 36 explains the de-jitter circuit 484 in more detail. Timestamp packets 480 may initially have some predetermined constant amount of time delay which is accounted for in line 500. For example, with no packet jitter, the timestamp packets 1-6 would be received at times intersecting line 500. However, due to packet jitter, the packets 1-6 can arrive at times that do not intersect line 500.
For example, a first timestamp packet # 1 may be received at time 504A and is longer than the expected receive time corresponding to time 504B on line 500. The timestamp de-jitter circuit 484 sends the expected timestamp value 504B to the timestamp rewrite circuit 494 that then replaces the received timestamp value 504A with timestamp value 504B. The timestamp de-jitter circuit 484 continues to identify the expected timestamp values for the received timestamp values. For example, timestamp # 3 is received at time 508A before it was expected at time 508B. The de-jitter buffer 484 sends timestamp value 508B to the timestamp rewrite circuitry 494 that then replaces the timestamp value 508A with timestamp value 508B.
The dejitter circuit 484 continuously tracks the actual received timestamp values. Over time, the received timestamp values may consistently be above or below the dejittered timestamp line 500. The dejitter circuit 484 may then move line 500 upward or downward according to the actually received timestamp value times. For example, over time timestamps may arrive at times 506, 510, etc. that are constantly above the current dejittered timestamp line 500. The dejitter circuit 484 accordingly adjusts line 500 upward to line 502. Timestamp line 502 is then used for replacing the received timestamp values with dejittered values. For example, after readjusting to line 502, a next timestamp # 6 may be received at time 514A. The de-jitter circuit 484 accordingly replaces the timestamp value 514A with de-jittered timestamp value 514B.
Similarly, if the received timestamp values over time are consistently below line 500, the de-jitter circuit 484 adjusts line 500 downward. The new adjusted lower timestamp line is then used for replacing timestamp values.
The timestamp rewrite circuit 494 replaces the received timestamp values with the timestamp values identified by dejitter circuit 484. The new timestamp values are then added to the DOCSIS data stream 488 or the MPEG data stream from video processing circuit 486 as previously done in conventional CMTS systems. The de-jittered timestamp along with the DOCSIS or video data is then sent to QAM 496, up-converter 498 and then sent over the downstream path of HFC 34.
The system described above can use dedicated processor systems, micro controllers, programmable logic devices, or microprocessors that perform some or all of the operations. Some of the operations described above may be implemented in software and other operations may be implemented in hardware.
For the sake of convenience, the operations are described as various interconnected functional blocks or distinct software modules. This is not necessary, however, and there may be cases where these functional blocks or modules are equivalently aggregated into a single logic device, program or operation with unclear boundaries. In any event, the functional blocks and software modules or features of the flexible interface can be implemented by themselves, or in combination with other operations in either hardware or software.
Having described and illustrated the principles of the invention in a preferred embodiment thereof, it should be apparent that the invention may be modified in arrangement and detail without departing from such principles. We claim all modifications and variation coming within the spirit and scope of the following claims.
Claims (22)
1. A modular Cable Modem Termination System (CMTS), comprising:
a packet shelf including a Data Over Cable Service Interface Specification (DOCSIS) Media Access Controller (MAC) configured to frame received data with DOCSIS frames and mark the DOCSIS framed data with destination addresses corresponding to downstream cable modems;
a physical interface (PHY) shelf configured to communicate with the packet shelf over a packet switched network, the PHY shelf to receive the DOCSIS framed data from the packet shelf;
wherein the packet shelf and the PHY shelf are distributed across the packet switched network; and
a timing system configured to synchronize DOCSIS timing between the packet shelf and the PHY shelf by dynamically adjusting timestamp values included in the received DOCSIS framed data on an individual basis according to the amount of delay experienced during the corresponding DOCSIS frame's traversal of the packet switched network.
2. The modular CMTS of claim 1 , wherein the timing system includes a timestamp master configured to generate a timestamp value in either the DOCSIS MAC or in the PHY shelf and send the generated timestamp over the packet switched network to a timestamp slave.
3. The modular CMTS of claim 2 , wherein the timing system is configured to add a time delay associated with transmitting over the packet switched network to a MAP advance time used for scheduling DOCSIS transmissions over an upstream path of a cable plant.
4. The modular CMTS of claim 1 , further including a timestamp de-jitter circuit configured to compare de-jittered timestamp values with timestamp receive times and over time synchronize according to the timestamp receive times.
5. The modular CMTS of claim 1 , wherein the timing system includes:
a first timestamp master wired directly to the packet shelf and configured to supply a first timestamp to the DOCSIS MAC; and
a second timestamp master wired directly to the PHY shelf and configured to supply a second timestamp.
6. The modular CMTS of claim 5 , further including:
a first global positioning system (GPS) receiver coupled to the first timestamp master and configured to provide a reference clock for generating the first timestamp; and
a second GPS receiver coupled to the second timestamp master and configured to provide a reference clock for generating the second timestamp.
7. The modular CMTS of claim 5 , further including a clock frequency master that operates independently in each of the packet shelf and the PHY shelf.
8. The modular CMTS of claim 1 , wherein the DOCSIS MAC is configured to conduct a ranging operation to identify an amount of delay required for sending a DOCSIS SYNC message to the PHY shelf and then add the identified delay to a MAP advance time used for scheduling DOCSIS data transmissions.
9. The modular CMTS of claim 1 , wherein the timing system is further configured to:
send a measurement packet over the packet switched network between the DOCSIS MAC and the PHY shelf;
measure a packet network time required for the measurement packet to travel over the packet switched network between the DOCSIS MAC and the PHY shelf;
determine a MAP advance time according to the measured packet network time and a conventional MAP advance that takes into account time required for the DOCSIS framed data to travel over a constant delay network between the PHY shelf and the downstream cable modems; and
use the MAP advance time to determine when to schedule DOCSIS transmission for the downstream cable modems.
10. The modular CMTS of claim 1 , wherein the timing system is further configured to:
transfer a DOCSIS SYNC message containing a reference timestamp over the packet switched network; and
synchronize clocks operating independently on the packet shelf and the DOCSIS MAC according to the reference timestamp.
11. An apparatus, comprising:
a downstream physical interface (PHY) shelf configured to receive Data Over Cable Service Interface Specification (DOCSIS) framed data from a remote DOCSIS Media Access Controller (MAC) over a packet switched network; and
a timing shelf configured to synchronize DOCSIS timing between the downstream PHY shelf and the remote DOCSIS MAC by re-writing timestamp values included in the received DOCSIS framed data on an individual basis according to the amount of delay experienced during the corresponding DOCSIS frame's traversal of the packet switched network;
wherein the apparatus is configured to transmit the DOCSIS frames having the re-written timestamp values over a modulated channel to be received by a downstream cable modem.
12. The apparatus of claim 11 , wherein the timing shelf is further configured to:
receive a reference timestamp transmitted over the packet switched network; and
synchronize a local clock with a remote clock operating on the DOCSIS MAC according to the reference timestamp, wherein the local clock is used for re-writing the timestamp values.
13. The apparatus of claim 12 , wherein the reference timestamp is synchronized to the Global Positioning System (GPS).
14. A method, comprising:
receiving, at a physical interface (PHY) shelf, Data Over Cable Service Interface Specification (DOCSIS) framed data from a remote DOCSIS Media Access Controller (MAC) over a packet switched network;
measuring a time delay associated with transporting the DOCSIS framed data over the packet switched network;
identifying timestamp values included in the received DOCSIS framed data; and
synchronizing DOCSIS timing between the remote DOCSIS MAC and the PHY shelf according to the measured time delay, wherein said synchronization includes correcting the timestamp values on an individual basis according to the amount of delay experienced during the corresponding DOCSIS frame's traversal of the packet switched network.
15. The method of claim 14 , further comprising:
sending a measurement packet over the packet switched network between the DOCSIS MAC and the PHY shelf;
measuring a packet network time required for the measurement packet to travel over the packet switched network between the DOCSIS MAC and the PHY shelf;
determining a MAP advance time according to the measured packet network time and a conventional MAP advance that takes into account time required for the DOCSIS framed data to travel over a constant delay network between the PHY shelf and a plurality of cable modems; and
using the MAP advance time to determine when to schedule DOCSIS transmission for the cable modems.
16. A system, comprising:
an edge device operating as a gateway between a constant delay portion of a cable network and a variable delay portion of the cable network, the edge device configured to receive Data Over Cable Service Interface Specification (DOCSIS) framed data over the variable delay portion and to quadrature amplitude modulate the DOCSIS framed data for downstream transmission over RF channels extending through the constant delay portion and to downstream cable modems;
wherein the quadrature amplitude modulation is performed separately from DOCSIS framing;
a timing component configured to identify timestamp values included in the DOCSIS framed data; and
the timing component configured to modify the identified timestamp values on an individual basis according to the amount of delay experienced during the corresponding DOCSIS frame's traversal of the variable delay portion of the cable network.
17. The system of claim 16 , wherein the timing component includes a timestamp master configured to generate a timestamp and send the generated timestamp over the variable delay portion of the cable network to a timestamp slave.
18. The system of claim 16 , further including a timestamp de-jitter circuit configured to compare de-jittered timestamp values with timestamp receive times and over time synchronize a local clock to a remote clock according to the timestamp receive times.
19. The system of claim 16 , further comprising a clock frequency master that operates independently of a remote clock frequency master operating in a remotely located DOCSIS Media Access Controller (MAC).
20. The system of claim 16 , wherein the timing component is further configured to:
receive a reference timestamp transmitted over the variable delay portion of the cable network; and
synchronize a local clock with a remote clock operating on a DOCSIS MAC according to the reference timestamp, wherein the local clock is used for modifying the timestamp values.
21. The system of claim 16 , wherein the timing component is further configured to transfer a DOCSIS SYNC message containing a reference timestamp over the variable delay portion of the cable network.
22. The system of claim 16 , wherein the timing component is further configured to:
receive a DOCSIS SYNC message containing a reference timestamp over the variable delay portion of the cable network; and
synchronize a local clock according to the reference timestamp, wherein the local clock is used to modify the identified timestamp values.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US12/413,690 US8160093B2 (en) | 2004-05-25 | 2009-03-30 | Timing system for modular cable modem termination system |
Applications Claiming Priority (10)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US57450604P | 2004-05-25 | 2004-05-25 | |
US57487604P | 2004-05-26 | 2004-05-26 | |
US58273204P | 2004-06-22 | 2004-06-22 | |
US58863504P | 2004-07-16 | 2004-07-16 | |
US59050904P | 2004-07-23 | 2004-07-23 | |
US62231204P | 2004-10-25 | 2004-10-25 | |
US62449004P | 2004-11-01 | 2004-11-01 | |
US63599504P | 2004-12-13 | 2004-12-13 | |
US11/134,818 US7539208B2 (en) | 2004-05-25 | 2005-05-20 | Timing system for modular cable modem termination system |
US12/413,690 US8160093B2 (en) | 2004-05-25 | 2009-03-30 | Timing system for modular cable modem termination system |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/134,818 Continuation US7539208B2 (en) | 2004-05-25 | 2005-05-20 | Timing system for modular cable modem termination system |
Publications (2)
Publication Number | Publication Date |
---|---|
US20090185574A1 US20090185574A1 (en) | 2009-07-23 |
US8160093B2 true US8160093B2 (en) | 2012-04-17 |
Family
ID=36698570
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/134,818 Expired - Fee Related US7539208B2 (en) | 2004-05-25 | 2005-05-20 | Timing system for modular cable modem termination system |
US12/413,690 Active 2025-09-17 US8160093B2 (en) | 2004-05-25 | 2009-03-30 | Timing system for modular cable modem termination system |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/134,818 Expired - Fee Related US7539208B2 (en) | 2004-05-25 | 2005-05-20 | Timing system for modular cable modem termination system |
Country Status (1)
Country | Link |
---|---|
US (2) | US7539208B2 (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9237116B2 (en) | 2013-09-12 | 2016-01-12 | Cisco Technology, Inc. | Network system time domain re-stamping |
US20190245677A1 (en) * | 2018-02-05 | 2019-08-08 | Arris Enterprises Llc | Two-Stage IP De-Jitter Algorithm in a Multiplexer for a Group of Statistically Multiplexed Single Program Transport Streams |
Families Citing this family (81)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7639617B2 (en) * | 2001-06-27 | 2009-12-29 | Cisco Technology, Inc. | Upstream physical interface for modular cable modem termination system |
US7209442B1 (en) * | 2001-06-27 | 2007-04-24 | Cisco Technology, Inc. | Packet fiber node |
US7688828B2 (en) | 2001-06-27 | 2010-03-30 | Cisco Technology, Inc. | Downstream remote physical interface for modular cable modem termination system |
US7324515B1 (en) | 2002-03-27 | 2008-01-29 | Cisco Technology, Inc. | Proxy addressing scheme for cable networks |
US7782898B2 (en) * | 2003-02-04 | 2010-08-24 | Cisco Technology, Inc. | Wideband cable system |
US8102854B2 (en) | 2004-05-25 | 2012-01-24 | Cisco Technology, Inc. | Neighbor discovery proxy with distributed packet inspection scheme |
US7817553B2 (en) * | 2004-05-25 | 2010-10-19 | Cisco Technology, Inc. | Local area network services in a cable modem network |
US7864686B2 (en) | 2004-05-25 | 2011-01-04 | Cisco Technology, Inc. | Tunneling scheme for transporting information over a cable network |
US7720101B2 (en) * | 2004-05-25 | 2010-05-18 | Cisco Technology, Inc. | Wideband cable modem with narrowband circuitry |
US7646786B2 (en) * | 2004-05-25 | 2010-01-12 | Cisco Technology, Inc. | Neighbor discovery in cable networks |
US7532627B2 (en) * | 2004-05-25 | 2009-05-12 | Cisco Technology, Inc. | Wideband upstream protocol |
US8149833B2 (en) | 2004-05-25 | 2012-04-03 | Cisco Technology, Inc. | Wideband cable downstream protocol |
US7835274B2 (en) | 2004-05-25 | 2010-11-16 | Cisco Technology, Inc. | Wideband provisioning |
US7701938B1 (en) | 2004-12-13 | 2010-04-20 | Cisco Technology, Inc. | Advanced multicast support for cable |
US7630361B2 (en) * | 2005-05-20 | 2009-12-08 | Cisco Technology, Inc. | Method and apparatus for using data-over-cable applications and services in non-cable environments |
US8379721B2 (en) * | 2005-09-22 | 2013-02-19 | Qualcomm Incorported | Two pass rate control techniques for video coding using a min-max approach |
JP4649315B2 (en) * | 2005-11-02 | 2011-03-09 | キヤノン株式会社 | Communication apparatus and communication method |
US8064414B2 (en) | 2005-12-13 | 2011-11-22 | Qualcomm, Incorporated | Range extension techniques for a wireless local area network |
US7701951B2 (en) * | 2006-03-06 | 2010-04-20 | Cisco Technology, Inc. | Resource reservation and admission control for IP network |
US8204074B2 (en) * | 2006-06-07 | 2012-06-19 | Broadcom Corporation | Flexible MAC/PHY association |
US8255682B2 (en) * | 2006-07-27 | 2012-08-28 | Cisco Technology, Inc. | Early authentication in cable modem initialization |
US7957305B2 (en) * | 2006-08-16 | 2011-06-07 | Cisco Technology, Inc. | Hierarchical cable modem clone detection |
US7865727B2 (en) * | 2006-08-24 | 2011-01-04 | Cisco Technology, Inc. | Authentication for devices located in cable networks |
US7899024B2 (en) * | 2007-02-28 | 2011-03-01 | Intel Corporation | Method and apparatus to support VoIP calls in an IEEE 802.16 interface |
US8665902B2 (en) * | 2007-03-12 | 2014-03-04 | Broadcom Corporation | Method and system for reducing transceiver power via a variable symbol rate |
US7773594B2 (en) * | 2007-07-11 | 2010-08-10 | Cisco Technology, Inc. | Transferring DOCSIS frames using a label switching network |
KR20090008529A (en) * | 2007-07-18 | 2009-01-22 | 삼성전자주식회사 | Apparatus and method for selecting of qos in portable communication system |
KR100932265B1 (en) | 2007-10-17 | 2009-12-16 | 한국전자통신연구원 | Packet transmission method and apparatus |
US8565692B2 (en) * | 2007-10-30 | 2013-10-22 | Lantiq Deutschland Gmbh | System and method for providing a versatile RF and analog front-end for wireless and wired networks |
US8208497B2 (en) * | 2007-12-28 | 2012-06-26 | Qualcomm Incorporated | Time stamped packet data interface between a modem and an RF unit |
US9112632B2 (en) | 2008-01-25 | 2015-08-18 | Cisco Technology, Inc. | Supporting efficient and accurate sync/followup timestamps |
US7860044B2 (en) * | 2008-05-30 | 2010-12-28 | Alcatel-Lucent Usa Inc. | Methods and apparatus for providing synchronization in a multi-channel communication system |
US8121116B1 (en) * | 2008-05-30 | 2012-02-21 | Cisco Technology, Inc. | Intra channel video stream scheduling |
US8797854B2 (en) * | 2008-09-29 | 2014-08-05 | Cisco Technology, Inc. | Scheduling for RF over fiber optic cable [RFoG] |
US8089983B2 (en) * | 2009-01-09 | 2012-01-03 | Cisco Technology, Inc. | Reserving resources over a DOCSIS downstream external physical interface |
US8160098B1 (en) | 2009-01-14 | 2012-04-17 | Cisco Technology, Inc. | Dynamically allocating channel bandwidth between interfaces |
US8861546B2 (en) * | 2009-03-06 | 2014-10-14 | Cisco Technology, Inc. | Dynamically and fairly allocating RF channel bandwidth in a wideband cable system |
US8428063B2 (en) * | 2009-03-31 | 2013-04-23 | Comcast Cable Communications, Llc | Access network architecture having dissimilar access sub-networks |
US8094661B2 (en) * | 2009-03-31 | 2012-01-10 | Comcast Cable Communications, Llc | Subscriber access network architecture |
US8787182B2 (en) | 2009-05-18 | 2014-07-22 | Comcast Cable Holdings, Llc | Configuring network devices |
US8973065B2 (en) * | 2009-07-17 | 2015-03-03 | Cisco Technology, Inc. | IPTV off-path admission control signaling in the presence of network address translation |
GB2476116A (en) * | 2009-12-14 | 2011-06-15 | Dublin Inst Of Technology | Real-time VoIP transmission quality predictor and quality-driven de jitter buffer |
US8644706B2 (en) * | 2010-01-22 | 2014-02-04 | Gainspeed, Inc. | Distributed cable modem termination system with software reconfigurable MAC and PHY capability |
US9887855B2 (en) | 2010-01-22 | 2018-02-06 | Alcatel-Lucent Usa, Inc. | Virtual converged cable access platforms for HFC cable networks |
US8935739B1 (en) | 2010-01-22 | 2015-01-13 | Gainespeed, Inc. | Distributed CCAP cable modem termination system |
US9584869B2 (en) | 2010-01-22 | 2017-02-28 | Gainspeed, Inc. | Virtual CCAP cable modem termination system with software reconfigurable MAC |
US10117006B2 (en) | 2010-03-31 | 2018-10-30 | Comcast Cable Communications, Llc | Hybrid fiber coaxial node |
US9344319B1 (en) * | 2010-06-30 | 2016-05-17 | Arris Enterprises, Inc. | Edge-QAM and M-CMTS core timing lock |
US8948043B2 (en) * | 2010-09-02 | 2015-02-03 | At&T Intellectual Property I, L.P. | Method and apparatus for normalizing cellular communications network data |
AU2012207388B2 (en) | 2011-01-23 | 2016-05-19 | Alpha Technologies Services, Inc. | Uninterruptible power supplies for use in a distributed network |
EP2715956B1 (en) | 2011-05-27 | 2019-09-11 | Cisco Technology, Inc. | Precision timing in a data over cable service interface specification (docsis) system |
US9088818B2 (en) * | 2011-06-21 | 2015-07-21 | Harman International Industries, Incorporated | Adaptive media delay matching |
US9246701B2 (en) * | 2011-09-16 | 2016-01-26 | Cisco Technology, Inc. | Downstream device architecture and control |
GB201116521D0 (en) * | 2011-09-23 | 2011-11-09 | Imagination Tech Ltd | Method and apparatus for time synchronisation in wireless networks |
US20130125194A1 (en) * | 2011-11-15 | 2013-05-16 | Harmonic Inc. | Converged Cable Access Platform for Provision of Video and Data Services |
CN102624560B (en) * | 2012-03-12 | 2016-06-29 | 深圳市天威视讯股份有限公司 | A kind of distributed deployment, centerized fusion CATV broadband access system |
US9742495B2 (en) * | 2012-10-25 | 2017-08-22 | Arris Enterprises Llc | Timing correction for a DOCSIS Edge-Qam |
US20140185630A1 (en) * | 2013-01-03 | 2014-07-03 | Casa Systems, Inc. | Systems And Methods For Performing Fiber Node Splits Using Remote QAM Modulation And Demodulation |
US9363188B2 (en) * | 2013-03-15 | 2016-06-07 | Arris Enterprises, Inc. | Cable modem termination system control of cable modem queue length |
US20140362714A1 (en) * | 2013-06-10 | 2014-12-11 | Cheetah Technologies, L.P. | Network-selecting hybrid fiber coaxial monitor, system including same and method of employing |
WO2014200280A2 (en) | 2013-06-12 | 2014-12-18 | Lg Electronics Inc. | Apparatus for transmitting broadcast signals, apparatus for receiving broadcast signals, method for transmitting broadcast signals and method for receiving broadcast signals |
US10050801B2 (en) * | 2014-01-31 | 2018-08-14 | Comcast Cable Communications, Llc | Digital forward link in hybrid fiber-coax network |
US9979564B2 (en) * | 2014-03-20 | 2018-05-22 | Cox Communications, Inc. | Virtual customer networks and decomposition and virtualization of network communication layer functionality |
EP3148129A4 (en) * | 2014-06-26 | 2017-08-16 | Huawei Technologies Co., Ltd. | Method and device for controlling quality of service of software defined network |
US10091071B2 (en) * | 2015-04-01 | 2018-10-02 | Nokia Of America Corporation | Assigning QoS to cable service flows |
JP6523058B2 (en) * | 2015-06-05 | 2019-05-29 | 株式会社東芝 | Communication system and estimation method |
EP3347979A4 (en) | 2015-09-13 | 2019-01-16 | Alpha Technologies Inc. | Power control systems and methods |
US10381867B1 (en) | 2015-10-16 | 2019-08-13 | Alpha Technologeis Services, Inc. | Ferroresonant transformer systems and methods with selectable input and output voltages for use in uninterruptible power supplies |
CN107205058B (en) * | 2016-03-16 | 2021-06-04 | 中兴通讯股份有限公司 | Media intervention control MAC address processing method and device |
US10523528B2 (en) * | 2017-05-22 | 2019-12-31 | Cisco Technology, Inc. | Determination of quality of service of a network tunnel |
US10635122B2 (en) | 2017-07-14 | 2020-04-28 | Alpha Technologies Services, Inc. | Voltage regulated AC power supply systems and methods |
US10349097B2 (en) * | 2017-10-27 | 2019-07-09 | Mti Film, Llc | Metadata editor for multimedia delivery |
US11018790B2 (en) * | 2017-11-06 | 2021-05-25 | Arris Enterprises Llc | Timing adjustment for distributed network architecture |
US10530512B2 (en) | 2018-04-27 | 2020-01-07 | Arris Enterprises Llc | Timing synchronization over cable networks |
US11758240B2 (en) | 2018-09-13 | 2023-09-12 | Avago Technologies International Sales Pte. Limited | Integrated video and data system |
US11546072B2 (en) * | 2018-12-14 | 2023-01-03 | Arris Enterprises Llc | Systems and methods to improve holdover performance in R-PHY network architectures |
US10478753B1 (en) | 2018-12-20 | 2019-11-19 | CH International Equipment Ltd. | Apparatus and method for treatment of hydraulic fracturing fluid during hydraulic fracturing |
US11498019B2 (en) | 2018-12-20 | 2022-11-15 | Haven Technology Solutions Llc | Apparatus and method for gas-liquid separation of multi-phase fluid |
JP7331431B2 (en) * | 2019-04-22 | 2023-08-23 | 日本電信電話株式会社 | Packet processing device and packet processing method |
TWI797554B (en) * | 2021-02-05 | 2023-04-01 | 新唐科技股份有限公司 | System on chip and control method |
DE102022118267B3 (en) * | 2022-07-21 | 2024-01-25 | Rockwell Collins Deutschland Gmbh | Synchronized data network system and method for initializing and synchronizing same |
Citations (35)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5457678A (en) | 1993-05-28 | 1995-10-10 | Siemens Aktiengesellschaft | Method and circuit arrangement for the transmission of message packets according to the asynchronous transfer mode in a communication network |
US6233235B1 (en) | 1999-11-01 | 2001-05-15 | Motorola, Inc. | Packet telephony power management |
US6331987B1 (en) | 1998-05-27 | 2001-12-18 | 3Com Corporation | Method and system for bundling data in a data-over-cable system |
US20020023174A1 (en) | 2000-03-20 | 2002-02-21 | Garrett John W. | Service selection in a shared access network using dynamic host configuration protocol |
US20020062450A1 (en) | 1999-05-07 | 2002-05-23 | Brian Carlson | Methods, modems, and systems for blocking data transfers unless including predefined communications to provide access to a network |
WO2002042882A2 (en) | 2000-11-17 | 2002-05-30 | Motorola, Inc., A Corporation Of The State Of Delaware | Robust method and apparatus for processing cable modem configuration files |
US20020097733A1 (en) | 2001-01-24 | 2002-07-25 | Kanta Yamamoto | Packet transmitting apparatus |
US20030026230A1 (en) | 2001-08-02 | 2003-02-06 | Juan-Antonio Ibanez | Proxy duplicate address detection for dynamic address allocation |
US20030043802A1 (en) | 2001-08-31 | 2003-03-06 | Takeki Yazaki | Packet transferring method and apparatus that employs the same |
US20030101463A1 (en) | 2001-11-26 | 2003-05-29 | Adc Telecommunications, Inc. | Passive CMTS redundancy |
US20030140131A1 (en) | 2002-01-22 | 2003-07-24 | Lucent Technologies Inc. | Dynamic virtual private network system and methods |
US20030163341A1 (en) | 2002-02-26 | 2003-08-28 | International Business Machines Corporation | Apparatus and method of dynamically updating dynamic host configuration protocol (DHCP) options |
WO2004006503A1 (en) | 2002-07-08 | 2004-01-15 | Packetfront Sweden Ab | Dynamic port configuration of network equipment |
US20040071148A1 (en) | 2001-08-21 | 2004-04-15 | Hitachi, Ltd. | Information device, gateway device and control method |
US6751230B1 (en) | 1999-05-24 | 2004-06-15 | 3Com Corporation | Upstream channel multicast media access control (MAC) address method for data-over-cable systems |
US6807193B1 (en) | 2000-06-20 | 2004-10-19 | 3Com Corporation | Cable modem with dribble grant access system and method |
US6829250B2 (en) | 2000-08-10 | 2004-12-07 | Verizon Communications Inc. | Automatic programming of customer premises equipment for vertical services integration |
US20050010958A1 (en) * | 2002-07-08 | 2005-01-13 | Rakib Shlomo Selim | Upstream only linecard with front end multiplexer for CMTS |
US20050078699A1 (en) | 2003-10-10 | 2005-04-14 | Broadcom Corporation | System, method, and computer program product for utilizing proprietary communication parameters to improve channel efficiency in a DOCSIS-compliant broadband communication system |
US6930988B2 (en) | 2002-10-28 | 2005-08-16 | Nokia Corporation | Method and system for fast IP connectivity in a mobile network |
US20050220126A1 (en) | 2002-07-11 | 2005-10-06 | Thomson Licensing S.A. | Application level gateway and firewall rule set download validation |
US20050226257A1 (en) | 2004-03-30 | 2005-10-13 | Adc Broadband Access Systems, Inc. | Virtual local area network |
US6986157B1 (en) | 1998-12-21 | 2006-01-10 | 3Com Corporation | Method and system for dynamic service registration in a data-over-cable system |
US6993353B2 (en) | 2001-03-14 | 2006-01-31 | At&T Corp. | Cable data service method |
US7006500B1 (en) | 2002-04-26 | 2006-02-28 | Valo, Inc. | Methods, apparatuses and systems facilitating data transmission across bonded communications paths |
US7023871B2 (en) | 2003-05-28 | 2006-04-04 | Terayon Communication Systems, Inc. | Wideband DOCSIS on catv systems using port-trunking |
US20060098669A1 (en) | 1996-07-25 | 2006-05-11 | Frederick Enns | High-speed internet access system |
US7050419B2 (en) * | 2001-02-23 | 2006-05-23 | Terayon Communicaion Systems, Inc. | Head end receiver for digital data delivery systems using mixed mode SCDMA and TDMA multiplexing |
US7110398B2 (en) | 2001-01-12 | 2006-09-19 | Broadcom Corporation | Packet tag for support of remote network function/packet classification |
US7197052B1 (en) | 2001-06-11 | 2007-03-27 | Cisco Technology, Inc. | Technique for interfacing MAC and physical layers of access networks |
US7467227B1 (en) | 2002-12-31 | 2008-12-16 | At&T Corp. | System using policy filter decision to map data traffic to virtual networks for forwarding the traffic in a regional access network |
US7492763B1 (en) | 2004-07-16 | 2009-02-17 | Applied Micro Circuits Corporation | User-specified key creation from attributes independent of encapsulation type |
US7548558B2 (en) | 2002-11-15 | 2009-06-16 | Terayon Communications Systems, Inc. | Cable modem termination system with flexible addition of single upstreams or downstreams |
US7613209B1 (en) | 2004-03-30 | 2009-11-03 | Extreme Networks, Inc. | System and method for egress packet marking |
US7646786B2 (en) | 2004-05-25 | 2010-01-12 | Cisco Technology, Inc. | Neighbor discovery in cable networks |
Family Cites Families (119)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US137606A (en) * | 1873-04-08 | Improvement in grain-driers | ||
US131766A (en) * | 1872-10-01 | Improvement in drawing-frames | ||
US135777A (en) * | 1873-02-11 | Improvement in lubricators | ||
GB8727846D0 (en) * | 1987-11-27 | 1987-12-31 | British Telecomm | Optical communications network |
US5153763A (en) * | 1989-12-01 | 1992-10-06 | Scientific-Atlanta, Inc. | CATV distribution networks using light wave transmission lines |
US6334219B1 (en) * | 1994-09-26 | 2001-12-25 | Adc Telecommunications Inc. | Channel selection for a hybrid fiber coax network |
US5604735A (en) * | 1995-03-15 | 1997-02-18 | Finisar Corporation | High speed network switch |
US6418324B1 (en) * | 1995-06-01 | 2002-07-09 | Padcom, Incorporated | Apparatus and method for transparent wireless communication between a remote device and host system |
US5784597A (en) * | 1995-09-22 | 1998-07-21 | Hewlett-Packard Company | Communications network system including acknowledgement indicating successful receipt of request for reserved communication slots and start time for said reserved communication slots |
US5805602A (en) * | 1995-09-25 | 1998-09-08 | Bell Atlantic Network Services, Inc. | Network monitoring system for cell delay variation |
US5931954A (en) * | 1996-01-31 | 1999-08-03 | Kabushiki Kaisha Toshiba | I/O control apparatus having check recovery function |
US5933420A (en) * | 1996-04-30 | 1999-08-03 | 3Com Corporation | Method and apparatus for assigning spectrum of a wireless local area network |
US5918019A (en) * | 1996-07-29 | 1999-06-29 | Cisco Technology, Inc. | Virtual dial-up protocol for network communication |
US5724510A (en) * | 1996-09-06 | 1998-03-03 | Fluke Corporation | Method of configuring a valid IP address and detecting duplicate IP addresses in a local area network |
US6101180A (en) * | 1996-11-12 | 2000-08-08 | Starguide Digital Networks, Inc. | High bandwidth broadcast system having localized multicast access to broadcast content |
US6233246B1 (en) * | 1996-12-30 | 2001-05-15 | Compaq Computer Corporation | Network switch with statistics read accesses |
US5963557A (en) * | 1997-04-11 | 1999-10-05 | Eng; John W. | High capacity reservation multiple access network with multiple shared unidirectional paths |
US6078595A (en) * | 1997-08-28 | 2000-06-20 | Ascend Communications, Inc. | Timing synchronization and switchover in a network switch |
US6137793A (en) * | 1997-12-05 | 2000-10-24 | Com21, Inc. | Reverse path multiplexer for use in high speed data transmissions |
US6510162B1 (en) * | 1998-05-27 | 2003-01-21 | 3Com Corporation | System and method for managing channel usage in a data over cable system |
US6640248B1 (en) * | 1998-07-10 | 2003-10-28 | Malibu Networks, Inc. | Application-aware, quality of service (QoS) sensitive, media access control (MAC) layer |
US6463078B1 (en) * | 1998-07-22 | 2002-10-08 | Microsoft Corporation | Method for switching protocols transparently in multi-user applications |
US6023769A (en) * | 1998-09-17 | 2000-02-08 | Apple Computer, Inc. | Method and apparatus for synchronizing an imprecise time clock maintained by a computer system |
US6381214B1 (en) * | 1998-10-09 | 2002-04-30 | Texas Instruments Incorporated | Memory-efficient leaky bucket policer for traffic management of asynchronous transfer mode data communications |
DE69939781D1 (en) * | 1998-10-30 | 2008-12-04 | Broadcom Corp | CABLE MODEM SYSTEM |
US6438123B1 (en) | 1998-11-10 | 2002-08-20 | Cisco Technology, Inc. | Method and apparatus for supporting header suppression and multiple microflows in a network |
US6804251B1 (en) * | 1998-11-12 | 2004-10-12 | Broadcom Corporation | System and method for multiplexing data from multiple sources |
US6606352B2 (en) * | 1999-01-15 | 2003-08-12 | Broadcom Corporation | Method and apparatus for converting between byte lengths and burdened burst lengths in a high speed modem |
US6763032B1 (en) * | 1999-02-12 | 2004-07-13 | Broadcom Corporation | Cable modem system with sample and packet synchronization |
US6546017B1 (en) | 1999-03-05 | 2003-04-08 | Cisco Technology, Inc. | Technique for supporting tiers of traffic priority levels in a packet-switched network |
US6377987B1 (en) * | 1999-04-30 | 2002-04-23 | Cisco Technology, Inc. | Mechanism for determining actual physical topology of network based on gathered configuration information representing true neighboring devices |
CA2371795C (en) * | 1999-05-26 | 2012-02-07 | Bigband Networks, Inc. | Communication management system and method |
US7161945B1 (en) | 1999-08-30 | 2007-01-09 | Broadcom Corporation | Cable modem termination system |
US6819682B1 (en) * | 1999-09-03 | 2004-11-16 | Broadcom Corporation | System and method for the synchronization and distribution of telephony timing information in a cable modem network |
US6490727B1 (en) * | 1999-10-07 | 2002-12-03 | Harmonic, Inc. | Distributed termination system for two-way hybrid networks |
US6526070B1 (en) * | 1999-10-09 | 2003-02-25 | Conexant Systems, Inc. | Method and apparatus for upstream burst transmissions synchronization in cable modems |
US7065779B1 (en) | 1999-10-13 | 2006-06-20 | Cisco Technology, Inc. | Technique for synchronizing multiple access controllers at the head end of an access network |
US6693878B1 (en) * | 1999-10-15 | 2004-02-17 | Cisco Technology, Inc. | Technique and apparatus for using node ID as virtual private network (VPN) identifiers |
US7117526B1 (en) | 1999-10-22 | 2006-10-03 | Nomadix, Inc. | Method and apparatus for establishing dynamic tunnel access sessions in a communication network |
US6999414B2 (en) * | 1999-10-27 | 2006-02-14 | Broadcom Corporation | System and method for combining requests for data bandwidth by a data provider for transmission of data over an asynchronous communication medium |
US7675938B2 (en) * | 1999-12-13 | 2010-03-09 | Broadcom Corporation | Flexible upstream PHY burst profile parameters to improve performance of short bursts in impulse noise |
US6698022B1 (en) * | 1999-12-15 | 2004-02-24 | Fujitsu Limited | Timestamp-based timing recovery for cable modem media access controller |
AU2737901A (en) | 1999-12-23 | 2001-07-03 | Broadcom Corporation | Method for opening a proprietary mac protocol in a non-docsis modem compatibly with a docsis modem |
US7269159B1 (en) | 2000-01-04 | 2007-09-11 | At&T Corp. | Upstream bandwidth allocation for packet telephony in a shared-media packet-switched access network |
US6857132B1 (en) * | 2000-01-14 | 2005-02-15 | Terayon Communication Systems, Inc. | Head end multiplexer to select and transmit video-on-demand and other requested programs and services |
US6993016B1 (en) * | 2000-11-16 | 2006-01-31 | Juniper Networks, Inc. | Methods and apparatus for transmission of analog channels over digital packet networks |
US7149223B2 (en) * | 2000-03-06 | 2006-12-12 | Juniper Networks, Inc. | Enhanced fiber nodes with CMTS capability |
FR2808353B1 (en) * | 2000-04-28 | 2003-12-05 | Airsys Atm S A | REDUNDANT INPUT/OUTPUT MANAGEMENT DEVICE, IN PARTICULAR COMPUTER ROUTING |
US6853680B1 (en) * | 2000-05-10 | 2005-02-08 | Bigband Networks Bas, Inc. | System and process for embedded cable modem in a cable modem termination system to enable diagnostics and monitoring |
US6847635B1 (en) * | 2000-06-20 | 2005-01-25 | 3Com Corporation | Method to transmit silence compressed voice over IP efficiently in DOCSIS cable networks |
US7274679B2 (en) * | 2000-06-22 | 2007-09-25 | Mati Amit | Scalable virtual channel |
JP3846771B2 (en) * | 2000-06-26 | 2006-11-15 | 三菱電機株式会社 | Decoder and playback device |
US20020088003A1 (en) * | 2000-06-27 | 2002-07-04 | Dror Salee | MAC redundancy in cable network headend |
US6771606B1 (en) * | 2000-06-29 | 2004-08-03 | D-Link Corp. | Networking switching system on-line data unit monitoring control |
US6950399B1 (en) * | 2000-07-06 | 2005-09-27 | Matsushita Electric Industrial Co., Ltd. | System and associated method for scheduling transport of variable bit-rate data over a network |
US6697970B1 (en) * | 2000-07-14 | 2004-02-24 | Nortel Networks Limited | Generic fault management method and system |
JP3656526B2 (en) * | 2000-07-17 | 2005-06-08 | 株式会社日立製作所 | Wireless communication base station, wireless position measuring system, transmission timing measuring device, and position measuring center device |
KR100672400B1 (en) * | 2000-11-20 | 2007-01-23 | 엘지전자 주식회사 | Apparatus and method for configuration file downloading in cable modem |
KR100390427B1 (en) * | 2000-12-06 | 2003-07-07 | 엘지전자 주식회사 | MAC frame format and method for communication setting in cable network |
JP3480927B2 (en) * | 2000-12-07 | 2003-12-22 | 株式会社東芝 | Cable modem, cable modem system |
US7039049B1 (en) | 2000-12-22 | 2006-05-02 | 3Com Corporation | Method and apparatus for PPPoE bridging in a routing CMTS |
US20020144284A1 (en) * | 2000-12-26 | 2002-10-03 | Burroughs Robert Sidney | Reliability enhancement for cable modem service |
US7006535B2 (en) * | 2001-01-12 | 2006-02-28 | Broadcom Corporation | Method and system for providing time offset to minislot clock and count in headend devices |
EP1356653B1 (en) * | 2001-01-24 | 2011-07-20 | Broadcom Corporation | Method for processing multiple security policies applied to a data packet structure |
US7145887B1 (en) | 2001-02-23 | 2006-12-05 | 3Com Corporation | Communication of packet arrival times to cable modem termination system and uses thereof |
US6490296B2 (en) * | 2001-02-28 | 2002-12-03 | Symmetricom, Inc. | Multi-link segmentation and reassembly for bonding multiple PVC's in an inverse multiplexing arrangement |
US20020133618A1 (en) * | 2001-03-14 | 2002-09-19 | Desai Bhavesh N. | Tunneling system for a cable data service |
US6993050B2 (en) * | 2001-03-14 | 2006-01-31 | At&T Corp. | Transmit and receive system for cable data service |
US20020147978A1 (en) | 2001-04-04 | 2002-10-10 | Alex Dolgonos | Hybrid cable/wireless communications system |
US6901079B1 (en) * | 2001-04-19 | 2005-05-31 | Cisco Technology, Inc | Providing different quality of services (QOS) to different point-to-point sessions |
US20020161924A1 (en) * | 2001-04-30 | 2002-10-31 | Perrin Robert E. | Method and apparatus for routing data over a computer network |
US6901580B2 (en) * | 2001-06-22 | 2005-05-31 | Intel Corporation | Configuration parameter sequencing and sequencer |
US7209442B1 (en) * | 2001-06-27 | 2007-04-24 | Cisco Technology, Inc. | Packet fiber node |
US7139923B1 (en) | 2001-06-27 | 2006-11-21 | Cisco Technology, Inc. | Technique for synchronizing network devices in an access data network |
US7639617B2 (en) * | 2001-06-27 | 2009-12-29 | Cisco Technology, Inc. | Upstream physical interface for modular cable modem termination system |
US7688828B2 (en) * | 2001-06-27 | 2010-03-30 | Cisco Technology, Inc. | Downstream remote physical interface for modular cable modem termination system |
US20030014762A1 (en) | 2001-07-13 | 2003-01-16 | Conover Richard David | Subscriber internet interface system and apparatus |
US7007296B2 (en) * | 2001-08-29 | 2006-02-28 | Terayon Communications, Inc. | Active cable modem outside customer premises servicing multiple customer premises |
US6779050B2 (en) * | 2001-09-24 | 2004-08-17 | Broadcom Corporation | System and method for hardware based reassembly of a fragmented packet |
US7206321B1 (en) * | 2001-09-27 | 2007-04-17 | Conexant Systems, Inc. | Method and apparatus for flexibly filtering scheduling messages in a communication system |
US7379472B2 (en) | 2001-09-27 | 2008-05-27 | Broadcom Corporation | Hardware filtering of unsolicited grant service extended headers |
US7599369B2 (en) * | 2001-09-27 | 2009-10-06 | Broadcom Corporation | Apparatus and methods for hardware payload header suppression, expansion, and verification in a DOCSIS network |
US20030066087A1 (en) * | 2001-09-28 | 2003-04-03 | Sawyer Wilson E. | Digital transmission system having modulators remotely located from central media access control layer |
US6959042B1 (en) * | 2001-10-01 | 2005-10-25 | Cisco Technology, Inc. | Methods and apparatus for measuring compressed video signals and applications to statistical remultiplexing |
US7333513B2 (en) * | 2001-10-09 | 2008-02-19 | Broadcom Corporation | Method, system, and computer program product for synchronizing voice traffic with minimum latency |
US7023882B2 (en) | 2001-10-19 | 2006-04-04 | Scientific-Atlanta, Inc. | Interfacing at least one information stream with at least one modulator |
US6763019B2 (en) | 2002-03-05 | 2004-07-13 | Nokia Corporation | Method and system for authenticated fast channel change of media provided over a DSL connection |
US7359332B2 (en) * | 2002-03-13 | 2008-04-15 | Broadcom Corporation | Enhanced DOCSIS upstream channel changes |
US7116643B2 (en) | 2002-04-30 | 2006-10-03 | Motorola, Inc. | Method and system for data in a collection and route discovery communication network |
US7206327B2 (en) * | 2002-05-17 | 2007-04-17 | Broadcom Corporation | Method and circuit for insertion of time stamp into real time data |
AU2003241616A1 (en) | 2002-05-24 | 2003-12-12 | Baker Hughes Incorporated | A method and apparatus for high speed communication with a downhole tool |
US20040248530A1 (en) | 2002-07-08 | 2004-12-09 | Rakib Selim Shlomo | Analog front end multiplexer for CMTS receiver |
JP2004064142A (en) * | 2002-07-24 | 2004-02-26 | Ntt Docomo Inc | Transmission power control method, radio communication system suitably adopting the same, radio base station, and mobile station |
US20040045037A1 (en) | 2002-08-27 | 2004-03-04 | Broadcom Corporation | Distributed cable modem termination system (CMTS) architecture implementing a media access control chip |
TWI227441B (en) * | 2002-10-11 | 2005-02-01 | Ambit Microsystems Corp | Firmware upgrade method for digital subscriber line device |
US7290046B1 (en) | 2002-11-22 | 2007-10-30 | At&T Corp. | System and method for detecting customer premise equipment behind a router on a data-over-cable system |
US7260165B2 (en) * | 2002-11-27 | 2007-08-21 | Broadcom Corporation | Method for synchronization through accelerated advance of counters |
US8995445B2 (en) * | 2002-12-19 | 2015-03-31 | International Business Machines Corporation | System and method for re-sequencing data packets on a per-flow basis |
US7782898B2 (en) * | 2003-02-04 | 2010-08-24 | Cisco Technology, Inc. | Wideband cable system |
US20040160945A1 (en) | 2003-02-13 | 2004-08-19 | Innomedia Pte Ltd. | Network communication system with a stand alone multi-media terminal adapter |
US20040181811A1 (en) * | 2003-03-13 | 2004-09-16 | Rakib Selim Shlomo | Thin DOCSIS in-band management for interactive HFC service delivery |
JP4381038B2 (en) | 2003-06-06 | 2009-12-09 | 株式会社東芝 | Transmission / reception device and cable modem module device |
US7363629B2 (en) * | 2003-06-19 | 2008-04-22 | International Business Machines Corporation | Method, system, and program for remote resource management |
EP1665765A4 (en) | 2003-08-29 | 2010-03-31 | Rgb Networks Inc | Video multiplexer system providing low-latency vcr-like effects and program changes |
US8320376B2 (en) * | 2003-09-05 | 2012-11-27 | Comcast Cable Holdings, Llc | Method and system for out-of-band messaging |
EP1685416A2 (en) * | 2003-11-17 | 2006-08-02 | General Instrument Corporation | Method and apparatuses for using packet data to manage a data stream in a broadband communications system |
US7849488B2 (en) * | 2003-12-23 | 2010-12-07 | Goback Tv, Inc. | Video modem termination system and method |
US20050198684A1 (en) * | 2004-03-08 | 2005-09-08 | Stone Christopher J. | Method and apparatus for providing a DSG to an OOB transcoder |
US7397825B2 (en) | 2004-03-10 | 2008-07-08 | Scientific-Atlanta, Inc. | Transport stream dejitterer |
US7809011B2 (en) | 2004-04-05 | 2010-10-05 | Broadcom Corporation | Method and apparatus for establishing communication between entities in a communication system |
US7372809B2 (en) * | 2004-05-18 | 2008-05-13 | Time Warner Cable, Inc. | Thwarting denial of service attacks originating in a DOCSIS-compliant cable network |
US20050289623A1 (en) * | 2004-05-21 | 2005-12-29 | Mowaffak Midani | Bulk tuning of frequency-modulated video signals |
US7835274B2 (en) * | 2004-05-25 | 2010-11-16 | Cisco Technology, Inc. | Wideband provisioning |
US7817553B2 (en) * | 2004-05-25 | 2010-10-19 | Cisco Technology, Inc. | Local area network services in a cable modem network |
US7720101B2 (en) * | 2004-05-25 | 2010-05-18 | Cisco Technology, Inc. | Wideband cable modem with narrowband circuitry |
US8149833B2 (en) * | 2004-05-25 | 2012-04-03 | Cisco Technology, Inc. | Wideband cable downstream protocol |
US7864686B2 (en) * | 2004-05-25 | 2011-01-04 | Cisco Technology, Inc. | Tunneling scheme for transporting information over a cable network |
US7532627B2 (en) * | 2004-05-25 | 2009-05-12 | Cisco Technology, Inc. | Wideband upstream protocol |
US7941512B2 (en) * | 2004-12-13 | 2011-05-10 | Cisco Technology, Inc. | Use of IPv6 in access networks |
-
2005
- 2005-05-20 US US11/134,818 patent/US7539208B2/en not_active Expired - Fee Related
-
2009
- 2009-03-30 US US12/413,690 patent/US8160093B2/en active Active
Patent Citations (36)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5457678A (en) | 1993-05-28 | 1995-10-10 | Siemens Aktiengesellschaft | Method and circuit arrangement for the transmission of message packets according to the asynchronous transfer mode in a communication network |
US20060098669A1 (en) | 1996-07-25 | 2006-05-11 | Frederick Enns | High-speed internet access system |
US6331987B1 (en) | 1998-05-27 | 2001-12-18 | 3Com Corporation | Method and system for bundling data in a data-over-cable system |
US6986157B1 (en) | 1998-12-21 | 2006-01-10 | 3Com Corporation | Method and system for dynamic service registration in a data-over-cable system |
US20020062450A1 (en) | 1999-05-07 | 2002-05-23 | Brian Carlson | Methods, modems, and systems for blocking data transfers unless including predefined communications to provide access to a network |
US6751230B1 (en) | 1999-05-24 | 2004-06-15 | 3Com Corporation | Upstream channel multicast media access control (MAC) address method for data-over-cable systems |
US6233235B1 (en) | 1999-11-01 | 2001-05-15 | Motorola, Inc. | Packet telephony power management |
US20020023174A1 (en) | 2000-03-20 | 2002-02-21 | Garrett John W. | Service selection in a shared access network using dynamic host configuration protocol |
US6807193B1 (en) | 2000-06-20 | 2004-10-19 | 3Com Corporation | Cable modem with dribble grant access system and method |
US6829250B2 (en) | 2000-08-10 | 2004-12-07 | Verizon Communications Inc. | Automatic programming of customer premises equipment for vertical services integration |
WO2002042882A2 (en) | 2000-11-17 | 2002-05-30 | Motorola, Inc., A Corporation Of The State Of Delaware | Robust method and apparatus for processing cable modem configuration files |
US7110398B2 (en) | 2001-01-12 | 2006-09-19 | Broadcom Corporation | Packet tag for support of remote network function/packet classification |
US20020097733A1 (en) | 2001-01-24 | 2002-07-25 | Kanta Yamamoto | Packet transmitting apparatus |
US7050419B2 (en) * | 2001-02-23 | 2006-05-23 | Terayon Communicaion Systems, Inc. | Head end receiver for digital data delivery systems using mixed mode SCDMA and TDMA multiplexing |
US6993353B2 (en) | 2001-03-14 | 2006-01-31 | At&T Corp. | Cable data service method |
US7197052B1 (en) | 2001-06-11 | 2007-03-27 | Cisco Technology, Inc. | Technique for interfacing MAC and physical layers of access networks |
US20030026230A1 (en) | 2001-08-02 | 2003-02-06 | Juan-Antonio Ibanez | Proxy duplicate address detection for dynamic address allocation |
US20040071148A1 (en) | 2001-08-21 | 2004-04-15 | Hitachi, Ltd. | Information device, gateway device and control method |
US20030043802A1 (en) | 2001-08-31 | 2003-03-06 | Takeki Yazaki | Packet transferring method and apparatus that employs the same |
US20030101463A1 (en) | 2001-11-26 | 2003-05-29 | Adc Telecommunications, Inc. | Passive CMTS redundancy |
US20030140131A1 (en) | 2002-01-22 | 2003-07-24 | Lucent Technologies Inc. | Dynamic virtual private network system and methods |
US20030163341A1 (en) | 2002-02-26 | 2003-08-28 | International Business Machines Corporation | Apparatus and method of dynamically updating dynamic host configuration protocol (DHCP) options |
US7006500B1 (en) | 2002-04-26 | 2006-02-28 | Valo, Inc. | Methods, apparatuses and systems facilitating data transmission across bonded communications paths |
US20050010958A1 (en) * | 2002-07-08 | 2005-01-13 | Rakib Shlomo Selim | Upstream only linecard with front end multiplexer for CMTS |
WO2004006503A1 (en) | 2002-07-08 | 2004-01-15 | Packetfront Sweden Ab | Dynamic port configuration of network equipment |
US7490345B2 (en) * | 2002-07-08 | 2009-02-10 | Terayon Communications Systems, Inc. | Upstream only linecard with front end multiplexer for CMTS |
US20050220126A1 (en) | 2002-07-11 | 2005-10-06 | Thomson Licensing S.A. | Application level gateway and firewall rule set download validation |
US6930988B2 (en) | 2002-10-28 | 2005-08-16 | Nokia Corporation | Method and system for fast IP connectivity in a mobile network |
US7548558B2 (en) | 2002-11-15 | 2009-06-16 | Terayon Communications Systems, Inc. | Cable modem termination system with flexible addition of single upstreams or downstreams |
US7467227B1 (en) | 2002-12-31 | 2008-12-16 | At&T Corp. | System using policy filter decision to map data traffic to virtual networks for forwarding the traffic in a regional access network |
US7023871B2 (en) | 2003-05-28 | 2006-04-04 | Terayon Communication Systems, Inc. | Wideband DOCSIS on catv systems using port-trunking |
US20050078699A1 (en) | 2003-10-10 | 2005-04-14 | Broadcom Corporation | System, method, and computer program product for utilizing proprietary communication parameters to improve channel efficiency in a DOCSIS-compliant broadband communication system |
US20050226257A1 (en) | 2004-03-30 | 2005-10-13 | Adc Broadband Access Systems, Inc. | Virtual local area network |
US7613209B1 (en) | 2004-03-30 | 2009-11-03 | Extreme Networks, Inc. | System and method for egress packet marking |
US7646786B2 (en) | 2004-05-25 | 2010-01-12 | Cisco Technology, Inc. | Neighbor discovery in cable networks |
US7492763B1 (en) | 2004-07-16 | 2009-02-17 | Applied Micro Circuits Corporation | User-specified key creation from attributes independent of encapsulation type |
Non-Patent Citations (3)
Title |
---|
cottage et al,DOCSIS 3.0 cable's position in the broadband market, BCI, 2005, 32 pages. * |
Fellows et al., "DOCSIS Cable Modem Technology", IEEE Communication Magazine, vol. 39, Issue 3, Mar. 2001, pp. 202-209. |
Stolowitz Ford Cowger LLP, Listing of Related Cases, Oct. 7, 2011. |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9237116B2 (en) | 2013-09-12 | 2016-01-12 | Cisco Technology, Inc. | Network system time domain re-stamping |
US20190245677A1 (en) * | 2018-02-05 | 2019-08-08 | Arris Enterprises Llc | Two-Stage IP De-Jitter Algorithm in a Multiplexer for a Group of Statistically Multiplexed Single Program Transport Streams |
US10951390B2 (en) * | 2018-02-05 | 2021-03-16 | Arris Enterprises Llc | Two-stage IP de-jitter algorithm in a multiplexer for a group of statistically multiplexed single program transport streams |
US20210211268A1 (en) * | 2018-02-05 | 2021-07-08 | Arris Enterprises Llc | Two-Stage IP De-Jitter Algorithm in a Multiplexer for a Group of Statistically Multiplexed Single Program Transport Streams |
US12021962B2 (en) * | 2018-02-05 | 2024-06-25 | Arris Enterprises Llc | Two-stage IP de-jitter algorithm in a multiplexer for a group of statistically multiplexed single program transport streams |
Also Published As
Publication number | Publication date |
---|---|
US20060168612A1 (en) | 2006-07-27 |
US20090185574A1 (en) | 2009-07-23 |
US7539208B2 (en) | 2009-05-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8160093B2 (en) | Timing system for modular cable modem termination system | |
US7688828B2 (en) | Downstream remote physical interface for modular cable modem termination system | |
US7639617B2 (en) | Upstream physical interface for modular cable modem termination system | |
EP2756636B1 (en) | Upstream external phy interface for data and control plane traffic | |
US10015052B2 (en) | Cross layer coordinated channel bonding | |
EP1965561B1 (en) | System and method for transmitting digital content using cable modem termination system (CMTS) bypass architecture | |
US7792034B2 (en) | Hierarchical flow-level multi-channel communication | |
US20090268751A1 (en) | Supporting Multiple Logical Channels In A Physical Interface | |
WO2018210169A1 (en) | Data transmission methods, devices, apparatuses, and system | |
KR101409924B1 (en) | Mixed serial and parallel stream channel bonding architecture | |
WO2005117310A1 (en) | Modular cable modem termination system | |
KR19990061605A (en) | How to Forward Synchronous Messages in Cable Modem Systems |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 8TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1552); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Year of fee payment: 8 |
|
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 12TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1553); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Year of fee payment: 12 |